(navigation image)
Home American Libraries | Canadian Libraries | Universal Library | Community Texts | Project Gutenberg | Children's Library | Biodiversity Heritage Library | Additional Collections
Search: Advanced Search
Anonymous User (login or join us)
Upload
See other formats

Full text of "Text and cases in marketing"

2Jthrart?B 






2Flje (gift of 

A. L. Kahl 



Digitized by the Internet Archive 
in 2013 



http://archive.org/details/textcasesinmarkeOOburs 



Text and Cases 
in Marketing: 

A SCIENTIFIC APPROACH 



Prentice-Hall International, Inc. 

London Tokyo Sydney Paris 

Prentice-Hall of Canada, Ltd. 

Prentice-Hall de Mexico, S.A. 



EDWARD C. BURSK 

EDITOR OF THE HARVARD BUSINESS REVIEW 



Text and Cases 
in Marketing: 

A SCIENTIFIC APPROACH 



PRENTICE-HALL, INC. 

ENGLEWOOD CLIFFS, N 
1 962 



Copyright © 1962 by PRENTICE-HALL, INC. 
Englewood Cliffs, N.J. 

All rights reserved. No part of this book 

may be reproduced, by mimeograph or any 

other means, without permission in 

writing from the publisher. 



Library of Congress Catalog Card Number: 

6 2-15413 



Printed in the United States of America 
90354-C 



Preface 



This book of text and cases is designed to meet today's conditions, where old- 
fashioned judgmental decision making must be supplanted by a more scientific 
approach. 

Marketing managers must now base their plans on more extensive and more 
objective data, and must evaluate their actions by more constant and more precise 
criteria, than ever before. Hence students of marketing must learn the basics of 
marketing research, and acquire the ability to use the information which that tool 
provides in analyzing and formulating marketing decisions. Conversely, marketing 
research cannot be performed fruitfully except by those with understanding of 
the marketing problems that require its help. Accordingly, this book is based on 
the premise that research and decision making can be learned best if they are 
studied as one integrated discipline in the classroom, just as, increasingly, they are 
used in actual practice. 

Also, in line with modern needs, this book contains material both in the 
behavioral field and in the statistical field. There should be no pretense that mar- 
keting has fully become a science, but it has made considerable progress toward 
that goal. Here only the rudiments of the scientific approach can be presented, 
although it is hoped that this will be enough to indicate its promise. Thus, for 
example, the text illustrates some basic uses of probability theory in decision mak- 
ing, and there are a number of cases to which this can be applied effectively; in 
general, however, the emphasis is more on scientific thinking than on scientific 
techniques. The keynote is systematic analysis and planning; even those who do 
not expect to engage in marketing may find here an intellectual discipline that 
will be useful to them in handling problems in other areas of activity and thought. 

The plan of this book— both in the way decision making and research are in- 
tegrated and in the role of marketing as the key element of dynamism in our 



VI PREFACE 

economy— has been developed in courses for American students, but has also been 
tested under even more rigorous conditions: it has been used for three years in 
the International Marketing Institute, where people from other countries are in- 
troduced to American marketing and have to learn it "from the ground up"— its 
content and, even more important, its spirit. Appreciation is due James A. Hagler, 
Executive Director, and Ernest D. Frawley, Treasurer of the Institute, for the 
opportunity to use material and for encouragement. 

Mr. Frawley, in particular, has been helpful in adapting research material for 
managers and in helping to put the book together, drawing on his experience as 
Project Director for the course in Marketing Research at the Harvard Business 
School. Edward E. Furash, and Stephen A. Greyser deserve special mention for 
their role as constructive critics; and Richard E. Armknecht, formerly Research 
Assistant, for his substantial contribution to the preparation of the cases. 

Above all, for unremitting toil and inexhaustible patience, I am grateful to 
Miss Virginia B. Fales, Managing Editor of the Harvard Business Review, and 
Miss Elizabeth H. Knox, chief typist and secretary. Many others also deserve 
thanks for help on many counts, particularly Miss Janet Risheill. Finally I want to 
acknowledge the intellectual stimulation of Professors Raymond A. Bauer and 
Robert O. Schlaifer, and the encouragement and understanding of Dean Stanley 
F. Teele. 

With the exception of 1-3, 2-2, 2-3, 2-4, 2-10, 3-6, 3-10, 4-1, 4-4, 4-9, 
5-1, 5-4, 5-5, 6-1, 6-8, and 6-9, all cases in this book have been individually copy- 
righted by the President and Fellows of Harvard College. They are reprinted 
here by special permission and may not be reproduced in whole or in part without 
the written permission of the President and Fellows of Harvard College. 

Edward C. Bursk 



Contents 



SECTION ONE 

THE NATURE OF MARKETING 1 

The Challenge 

Raising the Standard of Living 

The Complexity 

Scientific Approach 

Lack of Precision 
Planned Risk 

Systematic Analysis 

Model Building 
Logical Questioning 

Disciplined Creativeness 

Breaking the Shackles of the Past 
Foreign Application 
Time for Decisions 

Note on Cases 

SECTION TWO 

UNDERSTANDING THE CONSUMER 47 

Nature of Demand 

Consumer Behavior 

Psychological vs. Logical 
How Emotional? 
Social Influences 
Fashion 



vn 



Vlll CONTENTS 

Marketing Research 

Quantitative vs. Qualitative 
Finding Out Why 
Deeper Motivations 
Deciding on Research 
Difficulty of Implementation 
Marketing Mix 
Derived Demand 

Looking Ahead 



SECTION THREE 

PRODUCT POTENTIAL 129 c ^ 

Market Testing 

Experimental Design 
Proposition 
Comparison 
Planning 

Elements of the Test 

Selection of Test Media 

Source of Information 
Kind of People 
Size of Sample 
Sample Design 

Quota Sampling 

Research Plus 

SECTION FOUR 
EXPANDING DEMAND 229 

Action & Reaction 

Changing the Conditions 
Buying Motives 

Causing Change in Consumers 
Arena for Action 
Taking the Initiative 
Example: California Fruit Growers 
Rational Buying 



"' Selling Strategy 

Open Selling 
The Selling Job 



CONTENTS IX 



Sales Management 

Number of Salesmen 
Selection of Salesmen 
Sales Compensation 
Field Supervision 

Channels of Distribution 

Role of Intermediaries 
Strategy of Distribution 

Advertising 

Management Responsibility 
For the Sake of Selling 



SECTION FIVE 



COMPETITIVE POSITION 347 



Primary vs. Selective Demand 

Share of Market 

Taking Measurements 
Changes in Attitude 

Evaluation of Ads 

Economic Value of Research 

The Communication Element 
Giving More Meaning 

Role of Brand 

How Many Brands? 

Optimum Combination of Appeals 

The Question-Answer Process 

Creating 6* Maintaining Interest 

Casual or Buried Questions 

Exaggeration 

Experimental Games 

Role-Playing 

Indirect and Disguised Questions 

Spontaneous Reaction 

Structuring the Answers 
Verification 
Quantification 



X CONTENTS 

SECTION SIX 

STRATEGY AND PLANNING 469 

Product Strategy 

Timing of New Products 
Kind of New Products 
Market Segmentation 
Product Research Policy 

Pricing for Profit 

Level of Price 
Form of Prices 
Strategy of Competition 

Long-Range Planning 
Setting a Goal 
Kinds of Planning 
New Organizational Emphasis 

The Future 

NOTE ON SECONDARY SOURCES 565 



List of Cases 



SECTION ONE 



1. Brand A versus Brand B — A Mathematical Approach 23 

2. Diary of a Research Project in the Television Set Industry 31 

3. Decision at Boling Bros. Department Store 45 



SECTION TWO 



1. Ball Brothers 69 

2. Farmcrafter, Inc. 71 

3. Schmidt Packing Company 72 

4. Bender Mattress Company 75 

5. Prune Research (A) 79 

6. Prune Research (B) 87 

7. Dow Chemical Company (A) 96 

8. Bramton Company 108 

9. Southern Pine Paper Company 113 
10. New Era Specialty Paper Company 124 



SECTION THREE 



1. Pure Test Farms Dairy (A) 163 

2. Consumer Attitudes Toward Irradiated Food Products 166 

3. Spalding Market Research 169 

4. La Crosse Marketing Research Agency 173 

5. Pure Test Farms Dairy (B) 176 

6. Gem Appliance Company 130 

7. Lockheed Aircraft Corporation 186 

8. Dow Chemical Company (B) 193 

9. Seaborg Machine Tool Company 211 
10. Research for a New Magazine 215 



XI 



Xll 


LIST OF CASES 


SECTION 


FOUR 


1. 

2. 
3. 
4. 
5. 
6. 


Buying an Automobile 

Continental Appliance Company 

Hawaiian Pineapple Company 

Holiday House 

International Latex Corporation 

Readiness to Buy Certain Consumer Goods 


7. 

8. 

9. 

10. 


Towland Corporation 
Whiting Corporation 
Lightning Aircraft Company 
Analysis of Distribution Costs 



SECTION FIVE 



SECTION SIX 



9. The Rio Railroad 



281 
283 
295 
299 
306 
313 
322 
324 
330 
333 



1. Musselman Department Store 384 

2. Cantwell Packing Company (A) 386 

3. Cantwell Packing Company (B) 393 

4. Holden Company 397 

5. Better Soap Company 403 

6. Electra Oleomargarine 420 

7. The Carborundum Company (A) 426 

8. The Carborundum Company (B) 432 

9. Bay State Abrasive Products Company 447 
10. A & B Chemical Company 458 



1. Apex Chemical Company 503 

2. The Carborundum Company (C) 506 

3. Tidewater Mills, Inc. 513 

4. Schwartzbrau Brewing Company 518 

5. Dixie Instruments, Inc. 522 

6. Technique for Selecting New Products 531 

7. Latexet Rubber Company 54O 
The Grandholm Company 553 



554 



10. Company "X," '%" or "Z" 560 



SECTION ONE 



The Nature 
of Marketing 



If the ever-changing, growing, active function of marketing could be stripped to 

its bare bones, we would see that: 

Marketing covers all the operations in the process of causing goods to move 
from where they are manufactured or assembled to the hands of the ultimate 
consumer or user, except physical transportation. Included are marketing re- 
search, design or selection of products, decisions about wholesale distributors 
or retail outlets, personal selling, advertising or other promotion, pricing, and 
planning of over-all market strategy. 

At any stage in the process— manufacturing, wholesaling, or retailing — 
the prospective buyers are the market for the sellers at that level. The marketer 
attempts to sell them, that is, to persuade them to buy; and succeeds accord- 
ing to the financial resources they have at their command (including credit), 
their existing and potential desire to acquire, and the effectiveness of his 
marketing actions in inducing them to desire his product enough to spend 
their resources for it instead of for some other product or competing brand. 
But this definition is too bare, too cold, to do justice to the problems and 

opportunities that marketing deals with. Let us look at the opportunities, first, in 

order to understand the role that marketing fills in the dynamic American economy. 

THE CHALLENGE 

These opportunities result from the fact that, in a very real sense, ours is a 
marketing economy— not a production economy, no matter how resourceful and 

1 



Z THE NATURE OF MARKETING 

zealous we have been on the technical side, no matter how many teams of overseas 
visitors come to study the secret of our manufacturing productivity. It is the 
essence of a marketing economy to look forward, to perceive new opportunities 
in unsatisfied and often unrecognized needs, and then to take action to make these 
potential opportunities become real. 

Over the years, American businessmen have taken the initiative as promoters, 
sellers, merchandisers; and then production— mass production— has responded to 
take care of the demand provided by marketing, at a level of efficiency made pos- 
sible only by the available mass distribution. Take the case of Henry Ford: al- 
though he was a mechanical genius, the core of his achievement was the convic- 
tion (to others it seemed a dream) that there was a mass market for serviceable 
low-priced cars. 

Developing countries across the world are learning the same lesson: that to 
actually crank up an economy and make it spin takes more than money and 
machinery— it takes the drive and direction which marketing provides. Static and 
dynamic concepts of business make the difference. The static concept says, ''Here 
is a market; now how can we make the most money out of that market as it exists?" 
This might lead, for instance, to producing a small volume of goods at a high price 
for the few people who could afford the product in question. The dynamic con- 
cept, developed in American marketing and exemplified by businessmen like Ford, 
says, "Let's make these people want more than they ever dared to, so there will be 
enough of a market to make it possible to sell our product at a low price and in 
the long run sell enough to give us a greater return." 

The corollary of the dynamic concept is that, if many businessmen take the 
same approach, then, to achieve mass production, they also engage in mass em- 
ployment. As a result, buyers end up actually having more income to buy more 
things, and the standard of living will rise. In this sense, marketing of the creative 
kind— that is, marketing which does not accept the world as it is but operates on 
the principle that it can change the world (or a piece of the world)— is synonymous 
with delivering a higher standard of living. 1 

Raising the Standard of Living 

One of the significant problems ahead is the increasing population. Demog- 
raphers now predict three hundred million people in the United States by the end 
of this century, and four billion in the world. Estimates are continually rising, how- 
ever, because the urge to breed and beget is constantly enhanced as disease and 
poverty are beaten down and the standard of living increases. If anything, present 
population estimates are conservative. 

Marketing is responsible for many of the factors causing the population 
explosion. Why? Consider what has happened to the standard of living under the 
impulse of our distribution and marketing system. The standard of living— the 

1 See Paul Mazur, The Standards We Raise: The Dynamics of Consumption (New York, 
Harper & Brothers, 1953), passim. 



THE NATURE OF MARKETING 3 

bundle of goods and services that people can afford— not only is larger but con- 
tains a higher proportion of the "good things of life," the comforts and enjoy- 
ments of living above the level of mere physical subsistence. People work shorter 
hours, have more leisure, are better educated, live increasingly in suburban areas, 
have more laborsaving devices in their homes, worry less about jobs and supporting 
their families in conformance with their own criteria of what is proper, and gen- 
erally manage to enjoy their children more. So they are going to keep on having 
more children; there is no doubt about this trend. 

What does it all mean? Some people have assumed that an expanding popula- 
tion inevitably means prosperity— it doesn't. It could easily lower the standard of 
living because what we need, if increasing population is to mean increasing pros- 
perity, is an increasing production of goods and an increasing demand for goods 
in order to keep the production mechanism going. (For an explanatory example 
see Table 1-A; for another illustration turn to Figure 4- A, in Section Four.) 

Can our economy do this? Will it support this kind of population— fully sup- 
port it (even encourage it), or barely support it? These are the only conceivable 
alternatives. It is unthinkable that over a period of years our industrial produc- 
tivity, with the aid of increased mechanization and automation, cannot at least 
keep pace with our biological productivity. It is not a foregone conclusion, how- 
ever, because of the crucial role of marketing. Will the necessary marketing be 
provided? 

This is where the world picture becomes important again. We are trying to 
raise our standard of living, which means attempting to raise the demands of 
people for more and better things, as in the past. Now, however, we are trying 
to do this, not just for ourselves, but for the whole world— something that has 
never been attempted before in history. It will take the greatest kind of world-wide 
marketing drive: not just promotional campaigns, but the utilization of the whole 
concept and philosophy of marketing, of building demand and engendering new 
needs at the very time that old needs are being satisfied, thus stretching busi- 
ness to satisfy the ever-rising needs of people. 

One of the most interesting and most exciting things about marketing is that 
it, of all the functions of business, makes business work for itself. Most of the other 
functions in business take things as they are and devise very intelligent, very com- 
mendable ways to cope with problems as they occur. But marketing, in effect, 
deliberately creates new problems for the other functions of business to take care 
of. By changing people's wants, in the sense of bringing them to the fore quicker 
than they would have developed spontaneously, marketing actually pushes the 
pace of business so that the dynamic circle of industry keeps spinning faster and 
faster. It is the means by which a business lifts itself— or by which an industry, or a 
country, or, conceivably, the world lifts itself. 

The standard of living thus achieved in the United States— not just for a few 
but for the many— is illustrated in Table 1-B, based on a vast research project 
which Alfred Politz Research, Inc., did for Life magazine. Note, in particular, 



o 


o 


o 
o 
o 


o 


o 


o 


o 


o 


o 




o 


m 




m 


t> 


o 


" 




o 
o 

CO 



o o 

o o 

© ©^ 

O in 

in cm 

CM CO 



o o 
o o 
© o 



o o 

o o 

o. co 

o o~ 

o o 

o in 



o o 

o o 

© 4- © 

o o o 

o -^ o 

o t& cd 

o" ©~ 

o o 

CM O 



o o 

o o 

o cd 

cT o ©~ 

o in o 

© w- © 



o o o 

O in o 

O €«- ©^ 

©~ o" 

o o 

1-1 o 



o 
o 

CD 

©" o 
o in 

O €/3- 



o 

3 CD 

I S J5 

o bfi co 

c > o 

P < H 



o o o o 

o o o o 

©^ ©^ co ©^ 

cnT ©~ o~ o~ 

i-i o o o 

€«- ©^ ©^ ©^ 

©~ ©~ ©~ 

o o o 

oo eg o 

H ** CnT 



© o o o 

o o o o 

©^ o^ o^ ©^ 

©~ ©~ o" ©~ 

h o o o 

w- o^ ©^ ©^ 

o" ©~ ©~ 

in in o 

CM €/3- CO 



o o o o 
o o o o 
o o o o 



CM CM O O O O 

H H O O O O 

€«- €/* ©^ ©^ O^ ©^ 

©~ cT cT ©~ 

o o o o 

CM O 00 O 



o o o 
o o o 
o o o 



o o 

o o 

<=t ++ ©. - 

o o o o o o 

i-i CO o o o o 

€** €«- O^ ©^ ©^ O^ 

©~ ©~ ©~ ©~ 

o o o o 

© CO I> O 



o o 
o o 
o o 



o 


o 




O 


o 


o 




O 


C3 


©^ 




O 










in 


oo" 


O 


O 


CM 


&- 


tT 


o 


CO 




€«- 


©^ 

o" 
o 
©^ 

in 


* 








o 


© 






o 


o 






o 


o 














o 


00 






o 
in 


€/3- 




o 
o 








C3 




H 


O 


o~ 


u 


■^ 


O 






«9- 


©^ 


o 


o 


©~ 


o 


o 




o 


o 


"* 




o 










m 


co" 




■f 


CM 


€»- 




&■ 



© o 
o o 
o o 



o 


o 




O 


o 


o 




O 


<o 


o 




O 










©" 


00 


O 


O 


o 


39- 


■^ 


O 


in 




<*>- 


C3 

©~ 

O 
O 



o o 
o o 
o o 



o o 
o o 
o o 




o o o 

o o o 

O CD CD 

o~ ©~ ©~ 

o o o 

o o o 



o o 

o o 

o, ©^ 

o~ o" 

o o 

o o 



o o o 
in in o 

CM t- O 



o o o 

o o o 

©^ ©^ ©^ 

o~ o" ©~ 

o o o 

o ©^ o^ 

©~ ©~ o~ 

o in in 

O CM CM 



o o o 
o o o 
o o o 



o o 
o o 
o o 



o o o 
o o o 
o o o 



O CiJ 

> s 

u >, 
<D O 

* s 





ja 


*j 


V) 


£ 


<Sl 


cu 


a 


o 


0) 


-C 


T3 


* J 



SJ 



II 



CO 

. E 
•g I 

it 
ig 

cu 3 



THE NATURE OF MARKETING 5 

Notes to Table J -A 

A. Take Situation A as the starting point. (For purposes of simplification we assume that 
there are no manufacturing costs except wages and salaries, that all profits are paid 
out in dividends, and that all income is spent for units produced. See note below.) 

B. Illustrating the static or nonmarketing concept, population increases 25% but people 
have no more to spend, because production does not increase, because people aren't 
induced to buy more. The standard of living drops from 100 units per capita to 80. 

C. Alternatively, with same increase in population as in B, marketing foresees that de- 
mand can be increased at least as much as population; therefore steps up selling ef- 
forts, thus necessitating more production, which generates enough more income so 
people can pay for the added amount they buy. The standard of living remains 
constant. 

D. As a still more dynamic alternative, marketing, encouraged by the potential oppor- 
tunities it sees (if only prices could be reduced and a higher scale of operations 
achieved), increases its efforts still further than in C; engages in advertising; builds 
up demand to the point where production costs are cut and enough more units sold 
at the reduced prices to increase total revenue, thus once again generating the income 
to support the new higher standard of living — 160 units per person. Both profits and 
average wages are higher. 

E. Then, as the result of the higher standard of living, people are encouraged to have 
more children, and the population increases faster, thus giving marketers (if once 
again they follow the dynamic concept) the chance to build demand even faster 
and farther than population. In comparison with A, ability to purchase and supply 
of goods available have both quadrupled per person — so everybody can have four 
times as much. 

NOTE: This is of course highly simplified. Actually some of production costs would 
go for capital equipment, some of which would be indirectly returned as wages, profits, 
and dividends in the capital goods industries, and some of which would represent in- 
terest and would be returned indirectly through wages, profits, and dividends in the 
financial industry. Also, there needs to be some savings out of wages (to buy stock or 
enable banks and insurance companies to lend money to business firms) and/or some of 
earnings would have to be retained in the business. But the principles shown above re- 
main — and, in fact, are intensified, since the ratio between growing demand and growing 
production must itself grow to compensate for these diversions and lags. In short, the 
building of demand must proceed even faster and farther than shown above, and not 
only to keep the two sides of the demand-production equation even, but also to pre- 
serve the psychological climate where investors are willing to invest and businessmen are 
willing to increase their capital equipment (and promotional) expenditures. (It is for 
this reason that depressions are often as much psychological as economic — although 
psychological factors soon show themselves in the form of economic results.) 

that over half of U.S. families now have incomes of $4,000 or more per year; 
and that even those in the under $2,000 class, amounting to less than 18 per cent 
of the total, are still able to spend an average of $213 on automobiles and $97 on 
recreation. In other words, this is a practical demonstration of the fact that the 
average level of people's ability to buy "a standard of living'' is high— high enough, 
indeed, so that even in the case of those at the lower end of the scale the bundle 
of goods and services that they can buy has a large element of convenience and 
leisure-time items. 



THE COMPLEXITY 

Opportunities only exist because there are problems to solve. And the prob- 
lems in marketing result, in large part, from its complexity. 



TABLE 1-B. PERCENTAGE DIVISION OF TOTAL SPENDING AMONG MAJOR GOODS 
AND SERVICES IN THE UNITED STATES 

Household Income Before Taxes 





All 




$2,000 


$3,000 


$4,000 


$5,000 


$7,000 






House- 


Under 


to 


to 


to 


to 


to 


$10,000 




holds 


$2,000 


$3,000 


$4,000 


$5,000 


$7,000 


$10,000 


or over 


Total annual expendi- 


















tures per household 


$4,110 


$1,933 


$2,924 


$3,839 


$4,363 


$5,016 


$6,063 


$7,946 


Per cent of total ex- 


















penditures 


100% 


100% 


100% 


100% 


100% 


100% 


100% 


100% 


Food, Beverages, 


















Tobacco 


29 


36 


33 


30 


29 


28 


26 


24 


Clothing, Accessories 


12 


11 


11 


13 


12 


11 


13 


14 


Home Operation, Im- 


















provement 


19 


17 


20 


18 


19 


19 


18 


18 


Home Furnishings, 


















Equipment, Appli- 


















ances 


9 


7 


8 


8 


8 


9 


9 


10 


Medical, Personal Care 


5 


7 


5 


5 


5 


5 


5 


6 


Automotive 3 


14 


11 


13 


15 


14 


16 


15 


15 


Recreation 


5 


5 


5 


5 


6 


5 


5 


6 


Other b 


7 


6 


5 


6 


7 


7 


9 


7 


Total U.S. House- 


















holds (000's) 


49,140 


8,610 


7,080 


7,510 


9,250 


9,680 


4,680 


2,330 


Sample base 


10,243 


1,940 


1,546 


1,544 


1,882 


1,934 


913 


484 


Average number of 


















persons per 


















household 


3.3 


2.7 


3.2 


3.4 


3.4 


3.6 


3.7 


3.8 



a Expenditures on automobiles in this category refer to net outlays after trade-in allowances. 

b Includes life insurance premiums and non-medical professional services. 

Source: Richard A. Ostheimer, "Who Buys What?". Journal of Marketing, January 1958, p. 265. 



Marketing— that is, marketing performed intelligently and scientifically (as 
we shall define it presently)— is as difficult to do as it is important. For the busi- 
nessman who has the will and the wit to succeed, however, this is a very fortunate 
state of affairs; if he markets better, he has a valuable competitive edge. Oppor- 
tunities are usually commensurate with problems— the tougher the problem, the 
greater the reward for cracking it. 

There are a number of major reasons why marketing was, for a long time, re- 
garded as an art, which could be practiced only by men gifted with a happy blend 
of experience and intuition (called inspiration when it turned out right, bad 
judgment or worse when it turned out wrong). For the most part, these are the 
same reasons why marketing is now becoming a science— that is, yielding to an 
approach which makes the best possible use of logic and analysis. This is because 
the material is so intricate and intangible that hitherto it has not been tackled 
consciously and formally. The reasons are: 

(1) The buying-selling relation is a subtle, fluid interaction between two 
people— multiplied by the hundreds, or thousands— and those on both 
sides continually affect each other. 

(2) Those on the buying side are so many and so various, and so far out 
of the direct control of the seller, that the seller has to rely on persua- 



THE NATURE OF MARKETING / 

sion to lead them. Yet persuasion itself can never be 100 per cent ap- 
plicable to all the people— far from it. 

(3) The seller has at his command, and therefore must choose among, a 
whole array of possible ways and combinations of ways to affect the 
buyers he is trying to persuade. 

(4) The information at the seller's disposal is, at best, so imperfect that he 
can never be more than partly sure he knows how prospective ap- 
proaches will work— particularly since the market he is going to deal 
with will be changing as the result of his own efforts. 

(5) General conditions are continually changing, both in terms of economic 
forces and of social trends, so that even what is attractive when it is on 
the drawing board (for example, a new luxury automobile) may have lost 
its appeal by the time it reaches the market two or three years later 
because it no longer fits people's desire to spend or idea of what is 
fashionable. 

(6) Every seller has competitors who are always trying to outguess and 
outmaneuver him, often deliberately taking action just to spoil his 
carefully planned efforts. 

(7) Now that buyers, in the form of the public, live comfortably above the 
level of physical needs, competition is broadened. Every seller must 
contend not just with rival products of the same kind, but with rival 
uses for the same money (a new car versus a trip to Europe, a new suit 
versus a clock radio, and so on). 

(8) Differences between rival products and rival uses of money have become 
more difficult to discern and more psychological in nature, thus putting 
an extra premium on the penetration and persuasiveness of the selling; 
differences which otherwise would be just noticeable, or even not quite 
perceived, can be made stronger in buyers' eyes by the efforts of the 
marketer. 

(9) Because of the state of competition— with most companies invading 
other companies' "backyards" with "me-too" products, and being in- 
vaded in return— it has become more difficult for many individual firms 
to continue to grow. In the effort to reach certain customers more di- 
rectly or more effectively, the traditional channels of distribution have 
become scrambled; companies find themselves using mixtures of whole- 
salers, retailers, and so on— using some and bypassing others— to the ex- 
tent that they sometimes end up competing with some of their own 
customers. 

Or, conversely, in desperation, costs have been cut to the point where 
they impair selling drive. Selling effort has been diluted as it is spread 
over more products in the line, particularly products with less unique- 
ness or less clearly demonstrable superiority. Promotional expenditures 



8 THE NATURE OF MARKETING 

have increased. And, in general, margins have shrunk so there no longer 
is as much leeway in which to make marketing mistakes. 
(10) Investments in mechanization within the factory, and correspondingly 
in the promotion to move the products turned out, provide more oppor- 
tunity for small business to find a market among buyers dissatisfied with 
standardized products and services, at the same time that they make it 
more expensive in dollars and more dangerous competitively for large 
business to be off the target just a little in aiming for the mass market. 
Those ten reasons add up to the fact that in marketing problems today we 
deal with more subtle and complex variables, and that this calls for a more sys- 
tematic, analytical approach— in fact, for a scientific approach. (It should be 
noted here that the last three reasons apply just as much to industrial products 
as to consumer products— perhaps more so, because these competitive develop- 
ments are more recent, hence less recognized and prepared for, in the industrial 
products area.) 

But at the same time research techniques and theoretical concepts have 
become increasingly available, many of them developed in actual business use, 
and many of them borrowed from various academic disciplines such as sociology, 
psychology, and anthropology. So, while the difficulties push marketing in the 
direction of needing to be more scientific, there are also new tools which 
pull in the same direction by making the corresponding opportunities of being 
ahead of competitors more attractive for those who exert themselves to learn new 
ways, new disciplines, new attitudes, and new habits— such as are involved in the 
scientific approach. 



SCIENTIFIC APPROACH 

What is meant, specifically, by the scientific approach? "Scientific" implies 
three things in ascending order of importance: 

(1) Using scientific theories (and the techniques based on them) wherever 
available— including both the behavioral sciences and statistics; for ex- 
ample, the theory of cognitive dissonance in the communications and 
advertising field, and the theory of probability for purposes of sampling 
and decision making. 

(2) Using experimental designs to predict the outcome of marketing actions 
—particularly the manipulation of variables under controlled conditions 
(recognizing that this is much more difficult in the market than in the 
scientific laboratory). 

(3) Above all, using analysis to (a) understand past experience, (b) appraise 
present facts, and (c) decide on future action in a systematic, planned 
way (particularly including the development and use of models— even 
rough ones— where applicable). 



THE NATURE OF MARKETING 



Crucial to this approach is the necessity of always keeping the buyer in mind : 
the buyer who is the target— the moving target, which can shoot back! Let us 
consider some of the parameters in which this approach must work. 



Lack of Precision 

It must be emphasized that being scientific in marketing does not mean be- 
ing precise or certain. Consider the contrast with engineering: 

When an engineer gets ready to build a bridge, he makes some test 
borings, and figures what kind of pilings he must sink. In figuring solutions 
to his engineering problems, he has a handbook that tells him what the tensile 
strength is of structural steels, and he designs a bridge that he knows will 
reach from shore to shore and carry the projected load. 

But in marketing this cannot be done. What happens is that the test 
borings never provide data of 100 per cent accuracy. The information may be 
only 75 per cent correct, even if the best marketing research is used. In terms 
of odds, the chances are 75 per cent that the piling will hold, 25 per cent that 
it won't. Furthermore, when the marketer gets ready to build his bridge, the 
very fact that he puts his foundations down in a particular place changes 
the nature of the terrain he is dealing with. When he stretches the steel from 
here to there, the steel itself changes. The market reacts to his promotion, 
and his promotional tools gain or lose effectiveness in proportion to that 
reaction. 

So the difficult job in marketing, in the creative part of selling, is to 
consider what happens when the action taken itself changes the original con- 
ditions—conditions that, at best, were only 75 per cent correctly indicated 
by marketing research. Yet that 75 per cent figure still helps the marketer to 
do a better job. He needs it more than ever because he operates in an area 
where things act in this tricky, feedback kind of way. 



Planned Risk 

The point is that the businessman must build his bridges— his campaign to 
introduce products, his selection of distributors, his reduction or increase in price 
schedules, or whatever— even when the odds admit a chance of failure. In compe- 
tition, he could not afford to make the bridge 100 per cent safe even if he knew 
how; he would have to charge too high a toll, and lose traffic to other bridge 
entrepreneurs. His success comes in building more bridges that hold up than 
bridges that fall, so he comes out ahead; and this also means maintaining his ratio 
of successes to failures at least as high as his competitors' ratios, if not higher. 

This leads to the further point, which deserves strong emphasis: that being 



10 THE NATURE OF MARKETING 

careful does not necessarily mean being on the low or ''conservative" side in plan- 
ning. To illustrate: 

Suppose a new product is under consideration. The sales manager says, 
"Let's plan to manufacture a year's output of 600,000 of these lovely new 
things that our marketing research shows people will be crazy about." The 
production manager says, "Oh no, that would be too risky— we shouldn't buy 
all the new equipment needed— it is better to be on the safe side. Let's plan 
on 300,000 of these unproved widgets." 

What to do? It is not conservative to decide on 300,000, if the probable 
costs of being wrong on the low side are greater than the probable costs of 
being wrong on the high side. Conservatism— true conservatism— will go 
where the risks are least, and it can be the high side. For if the profit margin 
per unit is, say, $10, then if the engineer is wrong and the company could 
have sold 600,000 but made only 300,000, the costs are foregone profit to the 
extent of $3,000,000. While, if the salesmanager is wrong and it turns out 
that the company could sell only 300,000 but made 600,000, then the costs 
may be some small capital charges and expenses on inventory carryover, or 
stepped-up promotion to avoid the need to carry over— maybe $2 per unit or a 
total of $600,000, compared with $3,000,000. 

The trouble is, the foregone profit (or "opportunity cost") of $3,000,000 
never shows up on the books. Joe, the production manager, is patted on the 
back: "Joe, you figured it right on the nose— made 300,000, sold 300,000." 
But if the company produces 500,000, and Bill, the sales manager, is wrong by, 
say, 100,000, he gets criticized roundly for clogging up the warehouse and be- 
ing so foolish as to think the public would welcome a gimmick like that. Yet 
his estimate, under the assumed circumstances, is still several million dollars 
better than Joe's. 

The answer here— the decision, that is, because it is laying the company 
on the line— is not necessarily 300,000 or 600,000, or even the pat compro- 
mise halfway between, 450,000. The least risky course may be 500,000, and 
then 450,000 is wrong by about a half-million dollars. But whatever the exact 
figure, the point is that the manager should approach the decision on the 
basis of: risk is an integral part of marketing. 

Risk is almost as integral a part of marketing as it is of playing bridge 
or sailboat racing— you just don't gain as much as is possible unless you go 
down once in a while on a bid, or occasionally go over the starting line ahead 
of the gun. The only way to avoid such reversals is to bid so low that you 
never win a rubber, or start so far behind that you never win a race. Not that 
you try to make mistakes, but you plan to get as close to the dividing line 
between success and failure as maximizes your percentage of wins, at the risk 
—the planned risk— of once in a while having to pay some penalty. 



THE NATURE OF MARKETING 11 



SYSTEMATIC ANALYSIS 



Let us take a preliminary look at what was cited earlier as the most important 
element of the scientific approach— scientific analysis. In a very real sense this 
element embraces the other two (scientific theory and experimental design), since 
it is a prerequisite for their use; that is, it must take place before the marketer 
knows what techniques or designs are appropriate. Yet it can stand on its own, 
and must in some situations, because either the conditions are not suitable for the 
more specialized techniques, or the need is not great enough to justify the cost 
in terms of time or money. 

We were just looking at risks. The greater the risks, the more important it is 
to take them intelligently and systematically. Or, to put it another way, if you take 
risks intelligently and systematically, you can afford to take bigger ones for the 
sake of bigger potential gain. Thus, in the above situation, management could 
try to think through the probabilities of different sales volumes, as shown in 
Table 1-C. 

According to this computation, a decision to manufacture 600,000 is definitely 
a better risk than 300,000. More specifically, with the same odds and cost estimates 
the total profit on a decision to manufacture 600,000 would be $3,300,000 (com- 
pared with only $2,760,000 for 300,000). But 500,000 is even more attractive. 
Note that this is true despite the fact that the chances of actually reaching 500,000 
are only 30 out of 100 (that is, 25 per cent + 5 per cent = 30 per cent). The 
reason, of course, is that it costs only a little to be wrong on the high side, but it 
costs a lot (in fact, five times as much per unit) to be wrong on the low side; 
while the distribution of the odds is such that the chances of coming within, say, 
100,000 of 500,000 on the low side are much better than going over by 100,000 on 
the high side (30 per cent -j- 25 per cent + 5 per cent == 60 per cent, compared 
with 5 percent). 

Note that such figures do not indicate the marketing decision; they simply 
provide a framework for making the decision more intelligently. For example, it 
may very well be that management will reason that the figures for 400,000, 500,000, 
and 600,000 are so nearly alike— particularly in view of the necessarily rough na- 
ture of the estimates in the probability schedule— that reasons of competitive 
strategy should determine the figure to adopt. For example: 

Management might decide on 400,000, if judgment or past experience 

indicated that competitors were more likely to react aggressively to a larger 

volume of sales. 

Or a quite opposite line of reason might apply, such as: it is better for 

the sake of customer good will to have excess stocks for fast shipment; and, 

besides, we can sell the excess next year if we don't sell it this year, so let's 

go for 600,000. 



< 






CO 






W 






J 






PQ 






3 






O 






tf 






& 




^ 


fc 




§ 


O 




^ 


* 




» 


a 




"R 


> 




•a 


5 




--1 


5 




o 

CO 


9 




o 
a 


u 

w 




1 


Q 




e 


O 




■s 


g 




e 






S 


P 

H 
U 
< 


CO 


1 


S3 


■♦»» 




5 


8 

55 


<X 


§ 




1 


P 

S 


•ex 

! 




P< 


•♦^ 


is 


O 


s 


^2 


tn 


o 


<M 


o 


r-H 




2 


«* 


«* 


O 


co* 




>-h 


s 




H 


o 




< 


"*S 




H 
P 


1 




5 


o 











0> 

u 




f 




CP 


CM 








3 


<u 


"""* to 


43 "-* 




1 






Qi 


bo 


W " 


CO * 


•m u 


Mr, 


<3 






M 

43 




<D CO 
43 «* 

, | i-H 

s s 


< s 

l-H CO 


*2 

Sffl 


If A i 
than ( 
B[$10 






1—1 


o 


o u 














1— 1 

< 


fn 


u o 






• 






o 


n, 


<<-• M-H 


»-H 




CM 










o 


o 


o 


O 


o 


^ 


o 






o 


o 


o 


o 


o 


cs 






CD 


CD 


cs 


o 


CO 


c 




o 














$ 


o~ 






cT 


o 


© 


o 


o" 


?•> 


o 






CM 


o 


00 


o 


o 


ca 


CD 






»-H 


to 


o 


CM 


CO 


s 












H 


,— I 




o 








€^ 










s 


















"** 


















"*3 


O 


o 




o 


o 


o 


o 


o 


^ 


O 


o 




o 


o 


o 


o 


o 


a 


O 


o 




o 


o 


o 


o 


o 


















o 


o 


o 




o 


o 


o 


o 


o 


t c 


o 






•<r 


in 


■"sr 


in 


in 


^ 


in 






i-H 


CD 


l-H 


CM 


CM 



►o 


o 


o 


o 


o 


o 


o 


o 


o 


O 


o 


o 


o 


o 


o 


o 


t-A 


o 


o 


o 


o 


o 


o 


o 


















CO 


o 


o 


o 


o 


o 


o 


o 


•ts 


o 


CM 


CD 


o 


o 


o 


o 


s 


•<* 






c- 


CM 


o 


CM 


S 










^ 


^ 




^ 




€»- 












*v 
















cy 


















o 


O 


o 


o 


o 


o 


o 


o 


o 


o 


o 


o 


o 


o 


a 


o 


o 


o 


o 


o 


o 


o 


s 
















o 


o 


o 


o 


o 


o 


o 


^ 


o 


C5 


00 


in 


o 


in 


m 


o 


<n 




i-H 


c- 


Oi 


t- 


i-H 


CO 
















s 




v± 












^ 
















Ctf 
















•+«i 
















£ 


o 


o 


o 


o 


o 


o 


o 


■^ 


o 


o 


o 


o 


o 


o 


o 


CO 


CD 


o 


o 


o 


o 


o 


co 


£ 
© 
















o 


o 


o 


o 


o 


o 


o' 


o 


"tf 


o 


o 


o 


o 


o 


_co 


CM 




CM 


m 


CD 


m 


1-4 


Cj 
















ca 




«* 












^3 
















ca 
















*■«» 
















r° 


o 


o 


o 


o 


o 


o 


o 


CO 


o 


o 


o 


o 


o 


o 


o 


CO 


CO 


o 


o 


o 


o 


o 


CO 


o 
















o" 


o 


o 


o 


o 


o 


o 


o 


in 


o 


m 


o 


m 


m 




1— 1 




H 


CM 


CO 


CM 





3 9 



co T3 

to o 



45 « 



s r o 



eP t^ c^ qP aRI 

o in o m m 
»-H cm co cm i 



c^ 
O 
O 



co O 

CO CJ 



^ s s o o ^ 
3 § a ° * § 



•te 



o o 

o o 

o o 

o o 

o o 

i-< CM 



o o 
o o 
o o 



o o o 
o o o 
^ m co 



<Z CJ 
CU 

^ 43 

a CD 



12 



THE NATURE OF MARKETING 13 

Or management might take into account the fact that it had other 
machinery available which could easily be diverted to producing this product 
if demand were on the high side, and thus drop down to the original decision 
of 500,000. 

If one could put these additional factors into quantitative terms, they also 
could be worked into the computation. Thus, if the loss of good will caused by 
inability to ship immediately were figured to be fifty cents, then formula #2 would 
be B [$10A— $1.50 (C — A)], and, if you recompute on this basis, the best risk 
does turn out to be a decision to manufacture 600,000 rather than 500,000. On this 
basis Bill was right, just by intuition— and lots of luck! 

Model Building 

What we have just used on Bill and Joe is a kind of model. In this book a 
model is considered to be anything which shows the relations and interrelations 
of cause and effect in operational situations. It can be a physical model, like an 
intricate hydraulic machine designed to simulate the currents and crosscurrents of 
advertising effects; or a diagram on paper, like a flow chart of buying influences 
in the case of industrial purchases; or a symbolic representation in mathematical 
form, like an equation showing the relation between selling costs and return on 
investment; or a list of categories, like the elements of the marketing mix used in 
the following section, which really provides a pattern of causal factors; or a con- 
cept, like the "break-even" idea of selling different quantities of units at differ- 
ent margins as alternative ways of covering the cost of development and tooling; 
or most basic of all— needed in every case before a more elaborate model can be 
applied— "mapping out the situation" by breaking it down into its causal elements, 
often by means of questions (as described presently under "Logical Questioning"). 
The point is that the marketer is guided in coming to a decision by being 
able (a) to take the model (whether he can see it with his eyes and actually 
manipulate it by pencil or by computer, or whether he envisions it and works it 
within his own mind), and (b) to trace out what will happen if he changes this 
or that variable (that is, what the result will be if he makes an action decision). 

Thus, models can be very useful in marketing, but it is important to realize 
that: 

Models can vary from being extremely precise to extremely rough-and- 
ready, and they are more likely to be the latter than the former; hence the 
predictions they turn out will often only be approximations and can only be 
counted on with less than full confidence. 

Models are usually generalized in the sense that they are constructed to 
cover the general run of situations, so great care is called for in applying a 
model to a particular situation, to make sure the fit is good. 

Models are effective only to the extent that the information or estimates 
put into them are sound. And the input depends either on the quality of 



14 THE NATURE OF MARKETING 

observation (including collection of data) or on judgment— both of which 
are almost sure to be less than perfect, if not actually far short. 

Models may entail more cost to achieve smaller refinements than is 
justified by the improvement thus achieved; but this does not preclude the 
use of any or all models— it simply suggests that for some purposes we should 
be satisfied with the rougher and more informal kind of models (usually of 
the logical variety) instead of the intricate, formal kind (usually of the 
mathematical variety). 

In this book, we shall follow the principle that, for the very reason that 
models, or approaches using the model kind of thinking, are important in scientific 
marketing, it is crucial to refine and improve the quality of the input of informa- 
tion and thought— and that such improvement is probably more valuable and 
more feasible for practical marketing managers than further improvement in the 
models themselves. 

Let it be emphasized that the more formally scientific techniques are not 
being deprecated here. In fact, many of them will be mentioned in the text sec- 
tions and in the cases. But for the purposes of gaining a general mastery of mar- 
keting, the surest course is to concentrate on (a) understanding the essential na- 
me of scientific techniques, rather than accumulating a detailed knowledge of 
them (new ones are proliferating all the time); and (b) actually developing skill 
in the practice of logical analysis, in the sense of using the rougher, more informal 
kind of model (which can feed into any and all of the more advanced techniques 
as they develop). 

After all, the two biggest needs are (1) to know when, where, and how to 
resort to specialized techniques; and (2) to be able to supply the judgment ma- 
terial for their input and to translate their output into management action. These 
two needs are what this book primarily seeks to meet. 



Logical Questioning 

Accordingly, it is important to understand this basic model for logical analysis, 
which has been tested again and again in actual work situations, and can be 
applied effectively to the discussion of cases. Here is the framework: 

( 1 ) The first step in mapping out the situation is to translate the decision 
that is to be made into terms of the underlying problem. Take a man who is 
trying to decide whether to ask a girl to marry him. His underlying prob- 
lem: is she the girl that I want to spend the rest of my life with? That is the 
problem he must solve before he makes his decision. (This begs the question 
of whether he ought to marry her at all, if he feels he is enough in doubt 
that he must puzzle it through. But, in business situations too, managers 
sometimes need to reassure themselves that the decision they want to make 
is the right decision— and sometimes they are glad they stopped to think.) 



THE NATURE OF MARKETING 15 

In marketing, by comparison, the decision might be, for example, whether 
to lower the price of a product or not. But the underlying problem would be 
whether enough more would be bought at the lower price to increase the 
net return. 

(2) The next step is to break the underlying problem into a series of 
component subproblems in the form of more detailed questions. Can she 
cook? Is she affectionate? Have we the same leisure interests? Can she man- 
age a budget? (There is no implication here that these are the right questions 
for every man to ask every girl; they are just to suggest the kind of ques- 
tions. Like a manager in a particular business situation, our man will need to 
use his judgment as to what particular questions he ought to ask.) 

In the case of the marketing decision, again, the subproblems might be: 
Will a lower income level be tapped so we will get new users? Will present 
buyers feel they can afford to use more? Will retailers give it more push? 

(3) There could be any number of further layers of subquestions or 
subanalysis. "Can she cook?" could be further analyzed in terms of her train- 
ing, her interest in food and homemaking, and her natural skills. "Will pres- 
ent buyers feel they can afford to use more?" could be further analyzed in 
terms of the basic motivations people have for this kind of product, the 
financial resources or credit terms available to buyers, and the competing 
products that might be more attractive than additional purchases of this one. 
Note that in every instance the subquestions refer to the causes that bring 
about the result, represented by the larger question from which they stem. 

(4) At every point, the crucialness of each question can be weighed, and 
a subdecision made as to whether to seek further information on that ques- 
tion by means of marketing research and, if so, on what scale and in what 
form. (In our marriage-decision case, of course, the man will have already 
depth-interviewed the girl, and perhaps other girls in the past that will serve 
for comparison. But, in the marketing case, we might find we need some new 
research into consumer motivations.) 

(5) Finally, of course, the subanswers are carried back to the questions 
on the level above them, and subanswers are formulated at that level, to be 
carried up to the next level, until a final answer to the underlying problem 
is reached and translated back into the decision. (For the sake of the story 
but not of the analogy, let us imagine our hero, after balanced thought, de- 
cides to "p°P the question"; and that the girl, being untrained in scientific 
marketing, says yes without thinking— or did she have it planned that way all 
along? ) 

But even when we leave romance for the reality of business, note, still, 
the large role of judgment. More judgment than ever is called for, because the 
analysis makes it clear that judgment is more crucial in more places. It en- 
ters into how wisely the decision is translated into the underlying problem; 
how perceptively the subquestions are formulated; how deeply the questioning 



16 THE NATURE OF MARKETING 

is carried; how prudently the need for further information is weighed; how 
cogently the subanswers are related and accumulated; and, finally, how wisely 
(again) the results are translated back into a decision to take action. 
This kind of approach is particularly useful in analyzing cases like 1-3, 2-8, 
4-1, 4-7, 5-1, 6-1, and 6-8. 



DISCIPLINED CREATIVENESS 

The application of the scientific method to an activity that requires a high 
degree of conceptual ability and the need to innovate (often daringly) results in 
what may be called "disciplined creativeness." 

Note that any kind of scientific approach involves considerable amounts of 
judgment; but whether it is quantifying the estimates of odds, in the earlier ex- 
ample of Joe and Bill, or following a pattern of analysis like the above, the proce- 
dure does make the exercise of judgment more systematic— so long as it doesn't 
go to the opposite extreme and lull management into forgetting the original role 
of judgment and taking the final answer as scientific truth. The answer is not 
scientific; the method is. Marketing research can help in arriving at sounder esti- 
mates; logic can help to analyze them meaningfully; and therefore the answer 
may be better, but that's all— though competitively even a little "betterness" may 
have a great deal of leverage. 

Note, too, that no matter how much "science" is used in reaching decisions, 
there remains a vital need for courage and creativeness. It takes more courage to 
take action when science has led you as far as it can, and then in effect it tells 
you, "Now you know how much you don't know." Or it takes more courage 
to know the risks and face them than to face them blindly or with the false confi- 
dence that comes from thinking one has all the answers. 

Above all, creativeness is still indispensable in scientific marketing, just as it 
is in any science. The fact that estimates and judgment and courage enter in is 
not because marketing isn't scientific, but because it still— and perhaps forever- 
is too filled with unknown facts, interacting variables, and spontaneous actions to 
be completely expressed and understood in rigorous, quantifiable ways. Even in the 
most scientific of laboratories, someone must conceive of the possibility of new 
relationships of chemicals or circuits or whatever is to be tested— something above 
and beyond previously established knowledge (though obviously it may be sug- 
gested by observation or past experience). 



Breaking the Shackles of the Past 

Similarly, in marketing, it is not enough to take facts as they are or seem to be. 
This is not just a matter or dreaming up things like new advertising appeals. No 
matter how useful such recognized kinds of creative activity may be, we must 
not overlook the fact that all of marketing is creative, and that the essence of 



THE NATURE OF MARKETING 17 

marketing is to try to change the world so it will buy more of this or that product. 
This is why it can be dynamic and scientific at the same time. 

Marketing is also creative in a more particular, yet still pervasive, sense. The 
essence of every aspect of marketing is to use the lessons of the past (facts of expe- 
rience) without taking them for granted as fixed for the future. And here it is 
harder to be scientific in marketing than in some laboratory. This is not just be- 
cause in marketing it is harder to maintain neat, controlled conditions, or because 
the properties of the materials one is working with refuse to stay fixed or neatly 
predictable. Even more, it is because the facts one uses as a basis for making 
decisions today to cause action in the future are the result of decisions made in 
the past— and, very often, made by the same person or group that must now 
make the new decisions. 

As a result, two things can happen: (a) having once made a decision for a 
particular reason or purpose, it is hard to rethink it, or be conscious of the need 
to question it; and (b) the facts seem to prove that it should be perpetuated— 
they do, because what they are is an exact reflection of the original decision. Here 
is an illustration of this phenomenon, particularly striking because, if one could 
factor it out, it would add up to millions of dollars of lost revenue, just because 
it involves a large industry with millions of prospective buyers: 

The sales departments of the telephone companies used to think rural 
areas had less potential for extension phones, so they never tried very hard to 
sell them in such areas. Therefore, it is not surprising that for a long time the 
rural areas did remain a sparse market for this convenience product. And when 
anybody re-examined the question, the fact that the ratio of extension phones 
to main phones was less in rural areas did look like proof that this was not a 
promising market. 

Extension phones were originally conceived as luxuries for urban dwell- 
ers. But it does not follow today, just because, physically, people can make 
themselves heard 5 miles or 500 or 5,000 miles away with one telephone in 
the home, that, psychologically, extensions are still luxuries rather than ne- 
cessities (necessary conveniences for today's minimum standard of decent liv- 
ing); or that farmers and their families would not want or need — yes, need 
—extensions as much as "city folk." Isn't the barn farther from the front 
hall than the upstairs bedroom in an urban home? Aren't business and social 
contacts important to all people, no matter where they live? 

Now, however, having realized that the low ratio of extensions to main 
phones in farms might mean only that nobody ever tried very hard to sell 
them to farmers, a number of the telephone companies have begun actively to 
develop that market— and with considerable success. 

Foreign Application 

A very recent variation on this story, in a foreign setting, concerns an Italian 
soap manufacturer who went to his bank for a loan to build a new soap factory. 



18 THE NATURE OF MARKETING 

"What about detergents, which have virtually ousted soap in the United States?" 
he was asked. And he answered, "Oh, they'll never get anywhere in this country; 
they're against the habits and customs of our housewives." Again, the mistake was 
to look at the facts of the present and not realize they simply reflected the actions 
of the past. In the past, no one had sold detergents in Italy— any more than they 
had in the United States before they started to try to sell them here, in the be- 
lief they could change prevailing habits and attitudes. 

Whatever happens to detergents in Italy, the fact is that just a few years 
ago the same situation prevailed in France, where many customs and attitudes 
are basically quite similar to those in Italy. Marketers are beginning to find out 
that the basic similarities of people across the world are greater than the apparent 
differences— though it can be argued that people in the United States are maybe 
a decade or a generation ahead in their timing. And the company that introduced 
detergents in France has made considerable headway in that country. 

There is a clue here for the soap company in Italy— and for all other companies 
there and everywhere— that one can gain some insight into the possibilities of 
creating change by observing what has happened in other countries just a few 
years further along in development. Thus, entrepreneurs considering introducing 
supermarkets in India can observe what happened when a supermarket was 
opened in Tel Aviv. (Israel has just lately undergone considerable development; 
see Case 3-6.) In Tel Aviv, almost everybody, except the entrepreneurs involved, 
said it couldn't be done, either. But it was, and very successfully. Further, it has 
caused prices of many commodities to come down drastically in the other stores 
all over the city, thus raising consumption and sales and stepping up the tempo 
of the whole retail economy. Another effect is a social one: the supermarket has 
democratized people; all social classes now mingle in the same store. 

This is not to suggest slavish adoption of what the United States or any other 
country has done. Indian supermarkets would have to be different because, for 
example, a mass of people can't support a large supermarket by driving there in 
their automobiles; they just don't have cars. Another difficulty is the deficiency of 
packaging. In fact, to take an institution like a supermarket and adapt it to a new 
culture requires the same kind of creativeness that gave it birth in the first place. 
It has been suggested, for example, that in India the supermarket could itself run 
buses to various parts of the city where it is located; and crude packaging like 
paper bags could be used at first— until the demand was clear, whereupon a new 
packaging industry might begin to develop, thus once again accelerating the 
economy. 

Time for Decisions 

The fact is that it takes a creative leap of the mind to see new possibilities, 
like the idea of using buses just mentioned. Even breaking down a problem into 
its component parts, as the first stage of analysis, requires imagination as to what 



THE NATURE OF MARKETING 19 

those parts may be. And if there is an art to marketing, as well as a science, it may 
very well be the "feeling" for when it is time to ask oneself critical questions- 
tough questions, mean questions, such as "How do you dare to take for granted 
you know what you're doing?" Obviously, a busy executive cannot afford to stop 
and engage in researching or soul-searching on every one of the hundreds of 
decisions he makes in a day. 

The big decisions may stand out clearly, though even here it is often difficult 
to look into the facts and make sure they are more than the reflection of past 
decisions (which may have been wrong, or only half-right, or even if wholly right 
in the past, not necessarily right now), and more than one's own unilateral logic 
(which may be different from the buyers'). But even routine decisions add up 
cumulatively to important policy, and the wise manager will discipline himself to 
stop and second-guess every so often. 

All this only serves to underline, once and for all, that marketing research and 
all the other analytical aspects of scientific marketing should not lull manage- 
ment into a sense of false security. (Even marketing research itself can suffer from 
over-refinement: computers will only do what one tells them to, and it is there- 
fore possible to miss relationship among data that could be caught— sensed— by a 
person going over a pile of questionnaires manually.) Furthermore, as it is hoped 
the following chapters will demonstrate, there often are occasions when rough- 
and-ready research (like picking up the telephone and asking one of your dealers 
a few relevant qustions) or common-sense reasoning (like trying to break a prob- 
lem down into its parts and examining their relations) are more to the point than 
elaborate methodology. 

The over-all concept is still scientific— trying to bring to bear on all problems 
the careful, conscious, systematic analysis which is appropriate to each of them; 
and this includes the same careful, conscious, systematic appraisal of how far it 
is sensible to go in research or analysis, in view of (a) the cost of money, time, or 
effort involved, and (b) the size or cruciality of the problem in question. 



NOTE ON CASES 

At the end of each section of text are a number of "cases." For those who are 
unfamiliar with the "case method," some general observations may be helpful at 
this point. 

For the most part, these cases are descriptions of business situations calling 
for a decision on future action or an appraisal of past action, or both. A few are 
concerned with exposition of a technique, but even here the objective is to stim- 
ulate discussion of how the technique can be used or what it means in terms of 
management performance. Further, these cases are all real, that is, based on prob- 
lems and events that actually took place, although in some instances they are 
disguised as to names of people, companies, and geographic locations. 



20 THE NATURE OF MARKETING 

For purposes of starting your thinking and getting you to explore as many 
areas as possible, questions are appended to each case. These are not meant to be 
limiting, and there may be many times when those particular questions never get 
raised in a discussion at all. Remember you must use your own imagination to find 
clues to the map of the situation (to be subsequently confirmed or reported by 
the test of logic). While many lists of categories, factors, or other kinds of con- 
siderations are provided in the following sections as suggestions of where to look 
for cause and effect, it is the essence of the whole dynamic nature of marketing 
that people— students or businessmen— continually look for new ways to analyze 
and understand and plan for action. 

Furthermore, you are privileged and indeed urged, not to disagree with facts, 
but to "read between the lines" and doubt the judgments and opinions of people 
portrayed or quoted in the cases. Above all, you will gain from the cases to the 
extent you analyze everything you can. 

Keep in mind that the objective is to learn by exploration, discussion, even 
argument (in as logical a framework as possible). There are no clear-cut solutions, 
no black-and-white answers. The actions of marketing managers or researchers 
described in the cases may or may not illustrate good handling of marketing prob- 
lems or assignments; that is for you to decide. 

Moreover, often you will want further information or guidance in how to attack 
a case. There is considerable justification for not giving it to you, in the fact that 
businessmen themselves often do not have such information or guidance either, 
and one of the things a marketer must learn is how to operate as effectively as 
possible in such imperfect situations. Then, too, even when more material would 
be available, subsequently, as a result of research or other sources of help, there 
are countless times when one must analyze a situation, and try, in effect, to arrive 
at a tentative decision, just to find out if it is worthwhile or desirable to go ahead 
to the point of doing further searching or preparation. 

No matter what else is involved in a case you should always ask yourself: 

(1) Is research needed or justified by the nature and magnitude of the par- 
ticular situation? 

(2) If so, what specific kind of information do I need; and in what specific 
ways do I expect to use it? 

(3) What is the right source of such help; and what must I (putting your- 
self in the position of the responsible party in the case) do to see that a 
good job is done (or has been done)? 

(4) Can I have confidence in the results (or in demonstration of the need for 
still further searching or preparation)? 

(5) Now, how can I translate this marketing information into marketing 

action? 

* * * 

The three cases that follow herewith are designed specifically to exemplify 
three different approaches to marketing situations. (Only the third is a typical 
business-situation case as described above.) 



THE NATURE OF MARKETING 



21 



The first is a mathematical approach, based on the assumption that, if we only 
knew enough, we could reduce marketing strategy to equations. The merit of 
trying to do this, even at the risk of oversimplification, is that it often is helpful to 
see the pattern of interrelationships— if only to make one aware of what to keep 
his eyes open for. 

This is pretty much the operations research approach, which largely makes use 
of mathematical models. A simple and very useful example is the model developed 
by the DuPont company to show the interlocking factors that determine return 
on investment (see Figure 1-D). It is worth noting that return on investment 
is being used in an increasing number of companies as a measure of corporate 
performance (rather than total dollar profits, or profits as a percentage of sales) 
since, among other things, it focuses attention on the optimum use of the com- 
pany's resources, which often makes it worthwhile to go after additional sales 
even at a higher promotional cost per unit. 



( I TURNOVER 



RETURN 

ON 

INVESTMENT 



SALES 



divided 

by 



TOTAL INVESTMENT 



< 



multiplied 

by 







INVENTORIES 
ACCOUNTS RECEIVABLE 
CASH 


WORKING 
CAPITAL 











plus 



PERMANENT 
INVESTMENT 



^ 



EARNINGS 
AS % OF 
SALES 









EARNINGS 



divided 

by 

1 SALES 



SALES 



COST OF. SALES 



MILL COST OF SALES 
SELLING EXPENSES 
FREIGHT AND DELIVERY 
ADMINISTRATIVE 



Source: William H. Newman, Administrative Action (New York, Prentice-Hall, 1953), p. 417. 
Figure 1-D. Relationship of Factors Making Up Total Return on Investment 



Business games also make use of this approach with varying degrees of com- 
plexity and with varying amounts of luck (the unforeseen, the unforeseeable) 
built into the procedures. Teams, representing competing companies, play against 
one another, invest "money" and "time" in plant, research, inventory, training, 
promotion, and so on, and keep revising their plans as the computer puts every- 
thing together according to unstated but predetermined (though not necessarily 
static) ratios and reports back how they are doing vis-a-vis their competitors. This 
can be good training, not in learning specifically what to do in a given situation, 



22 THE NATURE OF MARKETING 

but in increasing the intellectual ability to sense and respond to complex inter- 
relationships, so that one is more likely to make good decisions in real life. 

The second approach is more pedestrian. It recognizes the untidiness of real- 
ity. It serves to emphasize the difficulties of getting the material to put into the 
equations or to fill out the models. It also recognizes the fact that, the more 
difficult it is to be scientific, the more important it is to try to be; and that, in the 
words of the old song, "a half moon is better than no moon, any old time, any old 
place." The half moon will help. So will the half science— so long as one recog- 
nizes the degree of imperfection, hence goes ahead with the less-than-full confi- 
dence that a half-lighted course requires, if one is to avoid stumbling. 

Marketing research is put forward in this case as a sort of substitute for 
science— for pure science. But can't it be more than that? What the case de- 
scribes is the process of trying to come to conclusions from the observation of 
data— data on events that have already taken place— almost like an astronomer 
deducing the forces that regulate the planets and stars from photographic plates 
that reflect patterns many light-years distant, hence far in the past. But the chem- 
ist, in his laboratory experiments, or sometimes the marketer in his tests, can "pl a y" 
with nature, that is, see what happens when he himself changes one of the var- 
iables. And he can test new arrangements that he himself has conceived. 

In some kinds of management problems, such as inventory decisions (which 
bear on marketing because the costs of production runs, warehousing, and trans- 
portation need to be balanced against the estimated effect on customer good will, 
but which deal mostly with physical flows and time sequences), considerable use 
of simulation is being made. From a mass of observed data a detailed model 
involving all the principal variables is made and put on the computer. Then the 
researcher or manager can "make a decision," feed it into the computer, and see 
what happens. 

Presumably, something like this could have been done in Case 1-2— or other 
similar marketing problems involving many interrelated intangibles—// all the 
necessary observed data were ascertainable and in hand. 

But the point is that in most such situations the data are, in fact, too diffuse, 
too difficult to reach and collect, for anything like simulation. The main reason 
is that what happens to manufacturers' products depends on a multitude of little 
decisions, not just on the part of consumers, but on the part of many interme- 
diate marketers in offices and stores in every city and town of the nation. 

Case 1-3 is a prime example of just one of these multitudinous situations, where 
one individual in a retail store will determine what happens to a slice of the po- 
tential market for the product in question. It is a simple, straightforward business 
problem— if there is such a thing. At least it looks that way on the surface. How- 
ever, what is called for is a digging down beneath the surface. The approach is the 
one that must be depended on for countless recurring problems : the act of analyz- 
ing, without any elaborate equations or data to help— except the equations of 
common-sense logic and the data of everyday observed experience. 



THE NATURE OF MARKETING 



23 



CASE 1 -1 

Brand A Versus Brand B — 
A Mathematical Approach 



It is now clear that the same reasoning 
processes which have led to notable prog- 
ress in the physical sciences can be applied 
to marketing, and that marketing "laws" 
can be derived in the same manner as the 
laws of physics. Therefore, the way is 
open for marketing to become, more and 
more, a science. 2 

This does not mean that formulas will 
replace the purely creative functions of 
marketing. However, in many operations 
where trial-and-error efforts have hitherto 
been necessary, marketing men will have 
the benefit of generalized information and 
insights similar to those which guide the 
work of engineers. 

PRACTICAL SCIENCE 

Through modern mathematics, practi- 
cal theories are being developed to guide 
and expedite decisions on complex ques- 
tions. I will attempt to build one such 
theory here. But first we must set the 
stage by stating this basic principle: Prac- 
tical theory must begin with tangible mar- 
keting actions and end in visible market- 
ing results. In short, practical theory must 
deal with marketing "observables." 

The key word in this principle is "ob- 
servables." In order for marketing theory 
to be practical, it has to work with things 



2 This approach is described by Harlan D. 
Mills, President of American Science Com- 
plex, Inc. See "Marketing as a Science," 
Harvard Business Review, September-Octo- 
ber 1961, p. 137. 



which marketing men can affect and see. 
But theory does not have to be based on 
observables! Let us consider an example 
from the natural sciences. 



Example 

Physics is certainly a practical science. 
Its theories have produced such practical 
results as atomic bombs and space satel- 
lites. Yet the things which physicists 
worry about — such as electrons, protons, 
gravitation, cosmic rays, and so forth — 
are not observable. No physicist has ever 
seen an electron, or a proton, or gravita- 
tion. They are figments of the physicists' 
imagination. Nevertheless, they play a vital 
part in theories which are used to produce 
practical results. 

We would do well to investigate the dis- 
tinction between observables and nonob- 
servables a little more closely, because it is 
so important in understanding the nature 
of scientific theories. Imagine an atom 
smasher — a very large and complicated 
piece of apparatus. There are wires run- 
ning in all directions, and magnets here 
and there. There is a switch with positions 
"on" and "off." At another place in the 
apparatus is a photographic plate. 

The observables are the switch, the pho- 
tographic plate, and the physical configu- 
ration of the equipment, wires, magnets, 
and the like. But associated with the ap- 
paratus are a host of nonobservables : we 
imagine electricity running up and down 
the wires, although we do not actually see 



24 



THE NATURE OF MARKETING 



it; we picture electrons jumping from one 
point to another in vacuum tubes; and so 
on. Through very complicated equations 
and reasoning, we hypothesize what would 
happen if the switch were turned on. In 
other words, we reason through nonob- 
servable activities until finally we are able 
to predict that something observable will 
occur. 

What we predict is that a white line 
will appear on the photographic plate. 
Now we have connected an observable 
cause and an observable effect. The ob- 
servable cause is the switch being turned 
on, and the observable effect is the white 
line appearing on the photographic plate. 

Thus we see that in physics one can 
trace observable causes through nonob- 
servable activities to observable effects. 
Every science has this same character. Per- 
haps it will be helpful to have it sum- 
marized: Physics is a practical theory deal- 
ing with physical observables. However, 
its power is due to nonobservables such 
as electrons (nonobservable objects) and 
gravitation (nonobservable mechanisms), 
which link observable causes and effects 
through physical theory. 

Systems of Nonobservables 

The need for nonobservables is even 
greater in marketing than in physics. If 
developing a theory were simply a matter 
of relating observables to each other, with- 
out any "submerged logic," then we would 
have had it long ago. The real task facing 
marketers is to develop theoretical systems 
of nonobservables which link observable 
causes and effects in marketing activities. 
Such theory must help the executive to 
ask proper questions and must prove out 
when tested on simple problems. 

To illustrate what is meant by "proper 
questions," suppose that you are playing 
stud poker. On the third card, with no 
strength visible, the man on your right 



puts up a big bet. Then you have two 
alternatives: 

( 1 ) You could ask yourself, "Is this guy 
bluffing?" That is not the proper ques- 
tion! Barring great psychological deficien- 
cies or differences, the question will lead 
nowhere. It is not a matter of whether the 
guy is bluffing, but of what you should do. 
True, the answer to the bluffing question 
makes the behavior question obvious. But 
it is false reasoning to conclude that you 
must answer the bluffing question in order 
to answer the behavior question. 

(2) Or you could ask, "How shall I 
play?" That is the proper question. The 
theory of games answers the question of 
how to behave in poker; and the problem 
of bluffing becomes an automatic by-prod- 
uct. On the surface, the distinction be- 
tween the behavior and bluffing questions 
may seem small, but, in fact, this distinc- 
tion is the difference between scientific 
progress and lack of it. 

As noted above, a theory must give cor- 
rect answers when proved out on simple 
problems: A word of explanation may be 
helpful: In order to use gravitational theo- 
ry, we first require that it predict simple 
things like "apples fall to the ground." 
Only then do we turn it loose on calcu- 
lating satellite trajectories. This is another 
basic attribute of science: that sound sci- 
entific theory must account for known 
facts and relationships before it proceeds 
to the unknown. 

In the theory we are about to develop, 
we will try to make sure that we have our 
apples falling to the ground before we ven- 
ture off into satellite predictions. 

COMPETITIVE EQUILIBRIUM 

The most characteristic aspect of mar- 
keting is competition. To be sure, market- 
ing deals with people, with motivation 
and communication, and with the organi- 
zation of resources. But competition is the 



THE NATURE OF MARKETING 



25 



one dominant feature. Marketing people 
constantly battle other marketing people, 
who are free to use new strategies and in- 
genious tactics to try to achieve their own 
goals. 

In our economy, a company of any size 
has, by definition, already passed stiff 
hurdles simply to stay alive and grow. 
There is a natural selection as ruthless and 
comprehensive as any that Darwin ever 
envisioned in the natural world. It selects 
companies that compete best, and propa- 
gates the more effective marketing strate- 
gies through their survival. 

Competitive equilibrium is the concept 
that most effectively characterizes a mar- 
keting situation in its over-all form. This 
is a state of a market (usually in dynamic 
movement) in which each competitor is 
acting to maximize his own profits against 
all competing strategies. It is recognized 
— and this is fundamental — that market- 
ing people must think not only about what 
they would do if the world should hold 
still, but also about what their strategy 
should be in the event that competitors 
should react intelligently with counter- 
strategies. 



Tactics in Competition 

To illustrate this concept, let us proceed 
to a study of the competition between 
two brands in a certain market. In this 
example of competition we simplify mat- 
ters for the sake of arithmetic; in actual 
cases, of course, the details can be as com- 
plicated as you want. Thus: 

Table 1-E shows two brands in compe- 
tition and their respective balance sheets. 
Each line of the balance sheet is derived 
from the preceding one. We assume, in 
this market, that a given share of the mar- 
keting effort leads to a like share of the 
market. (Notice that Brand A in this ta- 
ble expends 25 per cent of the market- 



ing effort and gets back 25 per cent of the 
market.) We also assume that the manu- 
facturing costs in each case are 50 per cent 
of the sales volume. Finally, the profit is 
simply the difference between the volume 
and the costs of marketing effort and 
manufacturing. 

TABLE 1-E. AN INITIAL POSITION 

Brand 



Marketing effort 
Sales volume 
Manufacturing costs 
Profit 



$100 
500 
250 
150 



B 

1 300 

1,500 

750 

450 



Further, just to keep the example easy 
to follow, suppose that total dollar sales 
in the market are fixed. As for marketing 
effort, it can take any form that is appro- 
priate for getting additional business — 
advertising, extra salesmen, price cuts, or 
whatever. 

Now put yourself in the position of 
Brand A and ask, "How can we improve 
our profit (which is currently $150)?" 
Brand A has the choice of raising or lower- 
ing its marketing effort of $100. One rea- 
sonable possibility is that profits will go 
up if marketing effort is increased. Let us 
try out that strategy and see what actually 
happens. Table 1-F shows the results when 
Brand A raises its marketing effort to 
$200. 

TABLE 1-F. A COMPETITIVE 
ADJUSTMENT 

Brand 



Marketing effort 
Sales volume 
Manufacturing costs 
Profit 



$200 
800 
400 
200 



B 

$ 300 

1,200 

600 

300 



Notice that even though Brand B does 
not change its marketing effort, Brand A's 
shift causes a change in the rest of Brand 
B's balance sheet. Now Brand A has 40 



26 



THE NATURE OF MARKETING 



per cent of the marketing effort and hence 
40 per cent of the total market. The rest 
of the balance sheet is recomputed as be- 
fore. We find that Brand A's profit does, 
indeed, go up— from $150 to $200. 

Now what should Brand B do at this 
point to counter? 

Offhand, it would seem that Brand B 
should raise its own marketing effort in 
response to Brand A's move. A trial case 
will show, however, that Brand B's best 
action is to reduce its marketing effort. We 
can calculate exactly how much it should 
be reduced, but let us consider the matter 
even a step further. Suppose we set up a 
sequence of steps in which Brand A, then 
Brand B, then Brand A, and so on, make 
moves and countermoves, each one con- 
tinually improving its own profit position. 

Where will this sequence ultimate- 
ly lead? It will lead to our concept of 
competitive equilibrium. This competitive 
equilibrium can be calculated mathemati- 
cally. For the case we are studying, it 
comes out as shown in Table 1-G. At this 
point, neither Brand A nor Brand B can 
improve its profit position by changing the 
marketing effort. 



TABLE 1-G. TWO BRANDS IN 
COMPETITIVE EQUILIBRIUM 



Brand 



Marketing effort 
Sales volume 
Manufacturing costs 
Profit 



PRACTICAL THEORY 



A B 

$ 250 $ 250 

1,000 1,000 

500 500 

250 250 



Now that we have this concept of com- 
petitive equilibrium, what can we do with 
it? Let us look at one example of how the 
concept can be used to develop some prac- 
tical marketing theory. Specifically, we will 
use the concept of competitive equilibrium 



to develop a theory which shows how a 
company can translate a cost advantage 
into the greatest possible competitive ad- 
vantage. 

Let us suppose that the manufacturing 
department of Brand A suddenly learns to 
make the product more cheaply. As a re- 
sult, manufacturing costs are now only 40 
per cent of sales volume instead of 50 per 
cent. If the marketing effort remains the 
same, the new balance sheet found in 
Table 1-H obtains. 



TABLE l-H. BRAND A ACHIEVES 
MANUFACTURING COST REDUCTION 



Brand 



Marketing effort 
Sales volume 
Manufacturing costs 
Profit 



A B 

1 250 $ 250 

1,000 1,000 

400 500 

350 250 



At this point, the $100 saving in manu- 
facturing costs has been transferred into 
profit. However, this $100 represents a 
potential war chest for the marketing de- 
partment. What would happen if it were 
reinvested in marketing effort, instead of 
pocketed as extra profit? Or, alternatively, 
what would happen if some fraction of the 
war chest were put into marketing effort, 
with the remainder going to profit? How 
do you decide what to do in this case? 
What is the best thing to do? 

To answer this question, we need a 
practical rule. And, indeed, we can close 
in on one, as we proceed with our ex- 
ample. 

Notice that Brand B has no cause to 
change on the balance sheet above, be- 
cause it is maximizing its profit against 
Brand A's marketing effort of $250. It 
is Brand A that has failed to maximize its 
profit against Brand B's marketing effort 
of $250, for Brand A's manufacturing 
costs have gone down, and this destroys 
the previous conditions of maximization. 



THE NATURE OF MARKETING 



27 



When Brand A does introduce a new 
strategy (by reinvesting some part of its 
cost saving in marketing effort), then 
Brand B's position is altered and it is 
forced to launch counterstrategy. Through 
a sequence of strategies and counterstrate- 
gies, Brands A and B will eventually arrive 
at a new competitive equilibrium. This 
new position will reflect the discrepancies 
in their manufacturing efficiency. Table 
1-1 shows the new competitive equilibrium. 

TABLE l-I. NEW STATE OF 
COMPETITIVE EQUILIBRIUM 

Brand 



Marketing effort 
Sales volume 
Manufacturing costs 
Profit 



A B 

i 298 $248 

1,092 908 

436 454 

358 206 



Notice that Brand A has reinvested $48 
of the possible $100 in extra marketing 
effort, and that Brand B's best response is 
to pull back slightly, by $2. Now Brand A 
has about 55 per cent of the business in- 
stead of 50 per cent. 

Possibly more surprising, however, is 
what has happened to Brand A's profit. 
The reinvestment of $48 in marketing 
effort has returned only an additional $8 
in profit. This seems to be mighty little 
for the money. But look what has hap- 
pened to Brand B's profit. It has dropped 
from $250 to $206 as the result of Brand 
A's strategy. By contrast, the share of profit 
for Brand A has been increased consider- 
ably. 

This points up a rather interesting phe- 
nomenon — that a cost advantage can and 
should be parlayed into additional advan- 
tages in share of market and share of 
profit. Our theory did not set out to prove 
this, but most marketing men will agree 
with the conclusion. It would appear (to 
refer back to our earlier analogy) that we 
have some apples falling to the ground. 



Incidentally, in studying this problem — 
and following the sequence of a manufac- 
turing improvement and a marketing im- 
provement — one cannot help but note 
that manufacturing people are on the job 
to "make money" and marketing people 
are on the job to "clobber the competi- 
tion." The major effect Brand A achieved 
by reinvesting part of its cost saving in 
added marketing effort was to cut down 
Brand B's profit; Brand A's own gain, as 
we saw, was relatively small. 



GUESSWORK GONE 

Why was it best for Brand A to reinvest 
$48? Why not $75, or $25, or $100? 
Why exactly $48? Without a guiding rule, 
Brand A's management would have to 
guess at the proper amount. But now we 
have a way to eliminate guesswork. When 
the mathematical basis for the increase is 
worked out, a simple rule emerges. 

We can phrase this rule in the follow- 
ing terms: At competitive equilibrium, a 
cost reduction should be reinvested in 
marketing effort in proportion to the in- 
crease of the unit manufacturing margin. 

This rule of thumb covers exactly what 
is relevant in the question and what the 
quantitative relationships are. (Because of 
the rule's importance, it is worked out 
in some detail in the Appendix.) In addi- 
tion, by omission, it points up all the ir- 
relevant things which one might be 
tempted to consider in connection with 
the problem. This is an important virtue 
of the scientific approach: it makes it pos- 
sible to stop worrying about elements of a 
problem which are really inconsequential. 

In the example presented here, it is 
completely unnecessary for Brand A mana- 
gers to know Brand B's manufacturing or 
marketing costs. As a matter of fact, 
Brand A managers do not have to know 
anything about their competitor. The key 



28 



THE NATURE OF MARKETING 



matter is the increase of the unit manu- 
facturing margin, and the required action 
is to reinvest a proportionate increase in 
the marketing effort. 

This rule meets the requirements we 
originally set down; it depends only on 
marketing observables, and it connects 
possible marketing actions with visible 
marketing results. 

Thus we have seen that it is possible to 
apply scientific reasoning to a dynamic 
and competitive marketing situation. For 
that reason, we can be confident that mar- 
keting can increasingly become a science. 



APPENDIX — RULE FOR REINVESTING 
SAVINGS IN MANUFACTURING COSTS 

In the example given earlier, we assume 
a fixed total market in sales of $2,000, 
split between two brands, A and B. In 
addition, we assume that: 

( 1 ) Share of marketing effort equals 
,share of market. 

(2) Manufacturing costs are a constant 
percentage of sales for each brand, say, a 
and b. Using the values in our example, 
a is equal to 50 per cent or 0.5 at all 
times, and b is equal to 0.5 at the begin- 
ning and 0.4 later on. 

Let x and y be the marketing effort (in 
dollars) expended by A and B, and let P 
and Q be their profits. Then P and Q can 
be built up as consequences of strategies 



x and y in the choice of levels of market- 
ing efforts. The generalized balance sheets 
in Table 1-J are developed as in the pre- 
ceding Tables 1-E through 1-1. 

That is, we have expressions for the 
profits as: 

(1A) P = 2,000(1 - a)— x 



(IB) Q = 2,000(1 - b)- 



- y 



The problem of the marketers is now to 
pick x to maximize P and y to maximize 
Q. But B's choice of y affects A's profit 
P, and A's choice of x affects B's profit Q. 

For any fixed value of y, the values of 
P, as they depend on x, describe a dome- 
shaped curve, as shown in Figure 1-K. 
(We assume that B's marketing effort is 
fixed.) A's profit P has a maximum value 
atx°. 

It can be shown, by differentiating P 
with respect to x, that to get this maxi- 
mum value of x°, A must choose x to 
satisfy the following maximizing equation: 

(2A) 2,000(1 - a)y = (x + y) 2 

Similarly, it can also be shown, for any 
fixed value of x, that to maximize profits, 
B must choose y to satisfy the equation: 

(2B) 2,000(1 - b)x = (x + y) 2 

When equations 2a and 2b hold simul- 
taneously, A and B are maximizing their 



TABLE 1-J. GENERALIZED BALANCE SHEETS 



B 



Marketing effort 


X 


Sales volume 


x(2,000) 

x + y 


Manufacturing costs 


x(2,000)a 

x + y 


Profit 


x(2,000)(l - a) 



x + y 



y 
y(2,ooo) 

x + y 

y(2,000)b 

x + y 

y(2 > 000)(l - b) 

x + y 



THE NATURE OF MARKETING 



29 



profits against each other's strategies; they 
are at competitive equilibrium. By equat- 
ing these expressions, we find: 

2,000(1 - a)y = (x + y) 2 = 2,000(1 - b)x 

An immediate consequence of this is the 
following relationship : 



(3) 



1 - a 1 - b 



This equation states that the marketing 
efforts at competitive equilibrium are pro- 
portional to the unit manufacturing mar- 
gins (1-a and 1-b) of the two brands. 




MAXIMUM VALUE 



\ 



Figure 1-K. Relation of Profit to Marketing 
Effort for Brand A 

By further substituting equation 3 into 
equations 2a and 2b we find, after some 
algebra, that: 

(«*),. 2,ooo ±0£r$ 

(4B )y = 2,000 V'i 1 .-,/ 

These are the levels of marketing effort 
which achieve competitive equilibrium for 
the brands. 

Finally, we substitute these marketing 
efforts into the profit equations 1a and 1b 
to find, after some simplification, that: 

(BA)P- 2,000 (2 ( . 1 -. a) ; )2 



(5B) Q = 2,000 



(1 - b) 3 
(2 - a - b) 2 



Now let us apply these formulas to our 
case example. When a =.5 and b = .5, 
equations 4a and 4b become: 

(.5) 2 (.5) 



x = 2,000 



y = 2,000 



(2 - .5 - .5) 2 
(•5)(.5) 2 



250 



250 



(2 - .5 - .5) 2 
These are the marketing efforts used in 
Table 1-G. Similarly, when a = .4 and 
b= .5, equations 4a and 4b become: 

(•6) 2 (.5) 



x = 2,000 



2,000 



(2 - .4 - .5)* 
(■6)(.5) 2 



298 



248 



(2 - .4 - .5) z 

Note that these are the figures for market- 
ing effort used in Table 1-1. The remain- 
ders of the balance sheets at competitive 
equilibrium in the exhibit follow directly. 

Implications for Strategy 

In order to derive the rule of thumb, 
notice how the marketing effort x in equa- 
tion 4a depends on the unit manufactur- 
ing margin (1-a). The derivative or rate 
of change of x with respect to the manu- 
facturing margin is: 



3x 



3(1 - a) 



4,000(1 - a )(l - b) 2 
(2 - a - b) 3 



This equation can be rewritten, using 
equation 4a again, as: 



3x 



2(1 - b) 



3(1 



i) (2-a-b)(l-a) 



Now, when a and b approach equality 
(as one could expect in a competitive in- 
dustry), notice that: 



(6) 



3x 



3(1 - a) 1-a 



In words, this equation states that the 
percentage change in marketing effort at 
competitive equilibrium approaches the 
percentage change in the unit manufactur- 
ing margin. This is the rule of thumb de- 
scribed earlier for determining the most 



30 



THE NATURE OF MARKETING 



profitable follow-up reinvestment of a 
manufacturing cost reduction in market- 
ing effort. Note that this is a rule of 
thumb, not an exact relationship. 

Other relations are also immediately 
available from equations 5a and 5b. For 
example, if the profits of Brand A are 
divided by the profits of Brand B, the 
following equation results: 



(7) 



Q VI - b^ 



In words, this equation states that an ad- 
vantage in the ratio of unit manufacturing 
margin can be cubed in profits. Such a 
parlaying of small edges into large advan- 
tages bears out general experience in mar- 
keting competition. 



How helpful will this approach be for 
marketing managers? What is likely to be 
its biggest advantage? Its biggest defi- 
ciency? What kinds of marketing deci- 
sions does it not apply to? 



THE NATURE OF MARKETING 



31 



CASE 1 -2 

Diary of a Research Project 
in the Television Set Industry 



STAGE 1 

Spring 1954. I become intrigued with a 
problem and cast it in the mold of a re- 
searchable project. 

March 17. For over two years now, my 
consulting work has enabled me to study 
various aspects of the television set indus- 
try. 3 Time and again, I confront large and 
sudden shifts in the market standing of 
individual manufacturers. But whenever I 
question executives in the industry, per- 
sons connected with the industry trade 
association, and/or the trade press about 
the causes for these shifts, I receive contra- 
dictory and unconvincing answers. Maybe 
it would be worthwhile to investigate 
thoroughly the major shifts that have oc- 
curred. Possibly, in this way I could un- 
cover the factors that make for market 
success and failure in this industry. Per- 
haps, also, these factors might to some 
degree apply to other industries. 

Data on the TV set industry are rela- 
tively plentiful, and at least a few of the 
top executives in the industry should be 
willing to give me access to them. Pre- 
sumably the factors that cause these major 
shifts (and possibly many of the minor 
ones, too) can be isolated with substantial 
research effort. Of course, I am not likely 
to reach demonstrable and duplicable con- 
clusions; however, those I do reach should 



3 The researcher in this case is Alfred R. Ox- 
enfeldt, Professor of Marketing, Columbia 
University Graduate School of Business. See 
"Scientific Marketing: Ideal and Ordeal," 
Harvard Business Review, March- April 1961, 
P . 51. 



be fully as reliable as those found in most 
writings in the field of marketing. 

Actually, it should be fairly easy to learn 
why major changes occurred in the market 
position of individual companies. Many 
persons associated with the industry seem 
positive that they know the answers, even 
though they have not made any special 
efforts to research the situation. With con- 
scious efforts to dig out the facts, I should 
get fairly reliable answers. 

But where should I begin? 

It seems to me that the best sources of 
information are studies made by companies 
whose market position has deteriorated 
substantially. The large firms, at least, 
must have gathered all available informa- 
tion in an effort to run down the causes 
of their difficulties. It may well be that 
my main job will simply be to obtain the 
studies made by companies that lost mar- 
ket position and find ways of telling their 
stories without embarrassing either the 
companies or their executives. 

This project could be valuable in many 
ways. The literature dealing with business 
contains no tested method that can be 
employed to account for past sales suc- 
cesses and failures in specific industries. 
The methods currently used in and out 
of business are fundamentally journalistic 
in nature. It certainly would be helpful to 
businessmen if a reliable scientific method 
could be found. 

Then, too, this project should help to 
satisfy the curiosity of many persons asso- 
ciated with the television industry. More 



32 



THE NATURE OF MARKETING 



to the point, knowledge of the actions 
and conditions that caused past improve- 
ments and declines in the market position 
of individual firms should be helpful to 
many as they set about making decisions 
for the future. After all, what could be 
more helpful to a businessman trying 
to select proper marketing policies than 
knowledge of the measures that worked 
out well in the past? 

But this, of course, raises another ques- 
tion. On what basis do businessmen select 
policies? When they say (as they always 
do when you ask them) that they rely 
mainly on "business experience," what do 
they really mean? When pressed, they 
state that business experience is a reliable 
basis for decision and action when it repre- 
sents (1) direct knowledge about many 
actions that have been taken by businesses 
in the past, (2) familiarity with the ra- 
tionale underlying them, (3) awareness of 
what happened as a result, and (4) a con- 
scious or unconscious processing of this 
information into generalizations. Thus, 
those who defend decisions based on busi- 
ness experience describe this experience as 
the product of a vast body of evidence 
which has been analyzed and compressed 
into nonexplicit generalizations. 

The study that I have in mind, accord- 
ingly, could actually reproduce or formalize 
business experience. It would attempt to 
do, consciously and explicitly, for one in- 
dustry, what successful business executives 
in all industries claim to do "naturally" or 
"subconsciously." If I should find that 
generalizations could not be made about 
the actions that produced sales success in 
this one industry, even after using pains- 
taking and thorough research methods, 
some very significant questions would be 
raised about the nature and validity of a 
whole class of business decisions. 

April 28. It is clear to me now that the 
project will be valuable, but just what it 
will involve is not so clear. In order to 



explain the actions that increased and re- 
duced market shares in the TV set indus- 
try, one would have to analyze the condi- 
tions surrounding major shifts in the 
market penetration of at least several large 
manufacturers. After six weeks of investi- 
gation, here is what I think that I would 
have to do before I could hope to arrive at 
generally acceptable conclusions: 

( 1 ) I would have to examine very 
closely the conditions and actions which 
precede every major change in each com- 
pany's market share. 

(2) Talks would have to be arranged 
with persons within the firm affected, with 
the firm's main rivals, with some of its 
distributors, retailers, and servicemen, and 
with members of the trade press. 

(3) Consumer product-testing journals 
would have to be studied to establish a 
history of comparative brand ratings. 

I realize clearly that conclusions reached 
will necessarily take verbal (rather than 
quantitative) form and will rest heavily on 
the recollections and interpretations of 
others. These interpretations are not likely 
to be unanimous, but they should suggest 
hypotheses that could be checked against 
objective information. 

But, at the very minimum, the topic is 
so valuable that it is worth undertaking — 
no matter what it involves and however 
dubious the conclusions which would be 
reached. Even if the results are "nega- 
tive," it would be extremely valuable just 
to know how much one could hope to 
learn about the causes of business success 
in a major industry. 

STAGE 2 

Summer 1954. The initial research method 
is selected and data collection begins. 

May 2. I have delayed collecting data 
long enough. I cannot expect to anticipate 
or resolve the many problems I will face 
until I actually dig into the facts. 



THE NATURE OF MARKETING 



33 



At this point, there seem to be two 
main methods of attack possible on this 
problem : 

(1)1 can try to explain selected major 
market share shifts that have taken place 
by drawing heavily on the explanations 
and, hopefully, the analyses made by the 
companies most directly affected. 

(2) I can take those factors that are 
most commonly used to account for shifts 
in market shares (especially those which 
can be quantified) and see whether they 
change substantially when shifts in market 
position take place. 

As I see it, the most important factors 
are: product quality, price, advertising 
expenditures, and dealers' and distributors' 
margins. Anybody can get a pretty good 
approximation of product quality by using 
the ratings published by Consumers Union 
and Consumers' Research. Manufacturers' 
list prices should serve nicely (then, too, 
for some time, a trade magazine has pub- 
lished list prices of all models of television 
sets). To measure changes in advertising 
expenditures, a variety of fairly reliable 
sources such as the Publishers' Informa- 
tion Bureau and Printers' Ink can be used. 
For retailers' and distributors' margins, the 
manufacturers' suggested prices — the ones 
they generally print for distribution — can 
be used. 

There is no need to choose between 
these two approaches, for they reinforce 
one another. Certainly any serious re- 
searcher must use informed business opin- 
ion as well as available published informa- 
tion. Unfortunately, neither the opinions 
expressed nor the information available 
will prove to be fully reliable, but doesn't 
this give all the more reason to use both? 



STAGE 3 

Fall 1 954 and Winter 1955. Dark days; 
the primary source of expected informa- 



tion evaporates and the crudity of available 
data becomes apparent. 

September 6. There certainly are many 
more difficulties in carrying out the TV 
project than I — or anyone, for that matter 
— could have expected. All companies I 
have approached, including some of the 
largest, simply do not study exhaustively 
the causes of their dips in market position. 
All they try to do is spot individual mar- 
ket declines as soon as they can and then 
attempt to learn the source of the trouble 
and try to correct it. 

Virtually nothing they think or do, how- 
ever, is a matter of written record. More- 
over, the efforts they put into diagnosing 
troubles are very superficial compared to 
what I had expected. This discovery repre- 
sents a major setback, for I shall have to 
do far, far more work than I originally ex- 
pected. On the other hand, it is of con- 
siderable interest and importance to know 
how casual are the methods of operation 
employed by some of the nation's largest 
and most successful firms. They are espe- 
cially lax about reviewing what happened 
and why — possibly because they are so 
pressed by problems of the present and 
the immediate future. 

In addition, there are many more hur- 
dles than I had anticipated in getting "re- 
liable enough" data about relative quality, 
price, advertising expenditures, and mar- 
gins for each major producer. On qual- 
ity of set, it appears that the consumer 
product-rating agencies only compare one 
or two models in the whole line of sets 
offered by most manufacturers. Typically, 
manufacturers offer about 15 to 25 models; 
therefore, those rated by the testing agen- 
cies may not be representative of the full 
line. This is especially likely to be true 
since agencies mainly test "promotional" 
models which often have a different chas- 
sis from other models of the same brand. 
Also, the individual manufacturers may 
want to provide different qualities in TV 



34 



THE NATURE OF MARKETING 



set performance. One may seek depend- 
ability; another, high sensitivity; another, 
clarity of picture or quality of sound; 
the criteria of quality that are used by the 
product-testing organizations may not take 
account of the diverse desires of different 
types of consumers. 

On prices of sets, there are also many 
headaches. How can one compare the di- 
verse prices and price changes for 15 to 
35 models of each brand of set? Would 
an unweighted average of all the prices 
serve? Would a weighted average be dif- 
ferent — and, if different, significantly bet- 
ter? Can one find different brands that 
are comparable in quality so that they can 
be representative for purposes of compar- 
ing prices and price variations? Do all 
prices in a line of TV sets move together, 
or do increases occur in the prices of some 
models while the prices of others decline? 
I shall have to find answers to these ques- 
tions. 

December 14. It seems to me, at this 
point, that potential errors about the ad- 
vertising expenditures of individual com- 
panies arise primarily from two sources: 

( 1 ) No information is published about 
outlays for "talent'' and production costs. 

(2) Two large set producers also own 
large TV and radio networks. They are 
therefore able to advertise in ways that 
make their expenditures a poor index of 
their actual advertising efforts. Indeed, one 
of them has been able to get testimonials 
from some of the best-known people in 
show business for nothing more than 
token gifts or a mere "thank you." 

Perhaps the hardest facts to pin down 
reliably are the margins that distributors 
and dealers obtain. It is all very well to 
learn the margins they would get if the 
prices they charged were in line with 
manufacturers' recommendations or with 
"list" or "allowed" margins. Actually, 
there has been price cutting and discount- 
ing on TV sets in some markets from al- 



most the very beginning of the industry's 
life. How, except by examining actual in- 
voices, can one measure the margins ac- 
tually obtained on TV sets by retailers and 
by distributors? The retailers and distribu- 
tors probably do not even know the an- 
swer themselves. 

March 1. I've learned a few lessons from 
the work done to date. I'd better set them 
down before I forget them. Some things I 
have learned deal with difficulties experi- 
enced with research methods, but others 
bear on the nature of business decision 
making and "business experience." For 
example: 

(1) Many of the things that people in 
and around the TV industry talk about 
(such as quality of product, price and 
price changes, margins, and advertising 
effort) cannot be defined rigorously or 
even in a manner that would insure sub- 
stantial uniformity of usage. Actually, I 
suspect they are so difficult to measure 
that people necessarily discuss them only 
on an impressionistic basis. As a result, 
people using terms like these, and appar- 
ently agreeing with one another, may be 
thinking about quite different things. 

(2) Since the impressions of executives 
and members of the trade press about such 
vital matters as relative price, margins, ad- 
vertising effort, product quality, customer 
satisfaction, and so on, are not even clear 
concepts — let alone based on factual com- 
parisons — how reliable can their memories 
of these matters be? Their memories, cer- 
tainly, are not reinforced by documenta- 
tion, for records never even existed about 
these things. One measure of the reliability 
of executive memory — which apparently 
underlies "business experience" — is the 
diversity of memories about the same situ- 
ation. Although I have not studied this 
point exhaustively, considerable difference 
of memory seems to exist even within the 
same firm. Needless to say, the variances 
are very great among executives in differ- 



THE NATURE OF MARKETING 



35 



ent companies. In part, there is a "com- 
pany line" regarding what happened in 
the past that serves to create a moderate 
uniformity of "memory" within a firm. 

(3) An incredibly slight amount of 
effort has been devoted (by even the larg- 
est manufacturers) to systematic and ob- 
jective analysis of individual market situa- 
tions, even when these situations require 
prompt remedies. Of course, most com- 
panies are making enough money so that 
their difficulties do not represent threats to 
survival. Whatever the explanation (and 
it is my impression that they operate in 
the same way with respect to their less 
profitable major appliances), I have found 
surprisingly little top-management interest 
in careful analyses of the past. Executives 
seem to have the feeling that they can get 
at the bottom of almost any problem 
in the course of a half-hour discussion 
with the persons involved. 



STAGE 4 

Spring and Summer J 955. The diverse ex- 
perience of each company in individual 
markets becomes apparent, and effort to 
explain changes in national market shares 
is abandoned. 

May 31. Troubles, troubles, and more 
troubles! Every time I try to explain a sub- 
stantial rise or fall in one company's na- 
tional market position, I find that the 
company has lost ground in some markets 
while gaining in others. Most cases I have 
examined lead me to conclude that a 
change in a company's national market 
share seems to have little relation to what 
happens in individual local markets. 

I have worked out Table 1-L to illus- 
trate the point. It shows the market 
shares of a large TV set manufactur- 
er during two successive years between 
1955 and 1960 in geographically scattered 
cities. Note that market shares went up in 
about twice as many markets as they went 



down (there were 11 increases and 5 de- 
clines). The range of manufacturer's 
shares of these 16 cities was between 1.6 
per cent and 11.3 per cent in the first 
year, and 2.3 per cent and 13.1 per cent 
in the second. It would have been far 
greater if all cities in which the company 
sold were included. 

TABLE 1-L. MARKET SHARES OF A 
LARGE TV SET MANUFACTURER 



City 


First year 


Second year 


A 


7.4% 


6.6% 


B 


1.6 


2.9 


C 


3.4 


2.3 


D 


5.1 


8.0 


E 


4.7 


5.7 


F 


10.0 


10.6 


G 


11.3 


13.1 


H 


6.3 


6.7 


I 


8.9 


9.1 


J 


5.4 


6.2 


K 


6.4 


6.1 


L 


3.4 


3.8 


M 


5.4 


3.8 


N 


4.0 


5.8 


O 


3.3 


2.3 


P 


10.6 


10.7 



Note: These data were obtained in confidence 
and are therefore disguised to conceal the identity 
of the company and the actual years involved. 

Table 1-M shows the market shares of 
two very large TV set manufacturers for 
five years and in various named cities. To 
ensure anonymity, the market shares are 
reported as index numbers. Specifically, 
the share attained in 1954 is taken as 100, 
and the shares in subsequent years are re- 
ported as proportions of the 1954 mar- 
ket share. For example, if the company 
had 5 per cent of a market in 1954 and 6 
per cent in 1955, its shares would be re- 
ported as 100 in the first year and 120 in 
the second; and if in the third year its 
share of that market declined to 4 per 
cent, it would be reported as 80. (Recog- 
nize that the sales in units during these 
three years need not vary in the same 
proportions.) 



36 



THE NATURE OF MARKETING 



This second table reinforces the con- 
clusions drawn from the first table. It 
shows variations in market penetration 
over periods of time to be very pronounced 
in a fairly representative group of cities. It 
also indicates that movements are diverse 
in direction as well as in size. 

TABLE 1-M. MARKET SHARES OF TWO 

LARGE TV SET MANUFACTURERS, 
COMPANY A AND COMPANY B, 1954-1958 

(Expressed as index numbers, 1954 - 100) 



City 



1954 1955 1956 1957 1958 



Providence, R.I. 




Co. A 


100 79.5 98.7 89.1 79.5 


Co. B 


100 95.9 99.3 95.3 97.4 


Newark, N. J. 




Co. A 


100 68.6 42.9 45.7 82.9 


Co. B 


100 100.6 98.3 108.2 104.9 


Washington, D. C. 




Co. A 


100 103.0 48.5 33.7 22.8 


Co. B 


100 100.4 90.2 84.4 71.3 


Charlotte, N. C. 




Co. A 


100 69.9 65.1 61.5 96.4 


Co. B 


100 119.4 123.1 134.2 135.2 


Miami, Fla. 




Co. A 


100 82.8 60.3 81.0 98.3 


Co. B 


100 110.2 113.2 116.9 105.1 


Dallas, Texas 




Co. A 


100 102.0 102.0 90.9 96.3 


Co. B 


100 111.2 109.7 95.8 100.8 


Memphis, Tenn. 




Co. A 


100 84.0 92.0 70.7 71.3 


Co. B 


100 115.6 102.3 122.3 117.2 


Cleveland, Ohio 




Co. A 


100 81.7 75.0 55.0 38.3 


Co. B 


100 111.6 104.2 103.1 101.8 


Huntington, W. Va 




Co. A 


100 115.0 126.0 78.4 85.1 


Co. B 


100 112.5 114.1 111.6 124.9 


Milwaukee, Wis. 




Co. A 


100 128.1 143.9 110.5 117.5 


Co. B 


100 125.3 132.3 134.2 125.9 


Louisville, Ky. 




Co. A 


100 100.0 85.8 74.2 75.8 


Co. B 


100 112.6 101.9 93.3 104.6 


St. Louis, Mo. 




Co. A 


100 81.1 68.9 60.0 68.9 


Co. B 


100 118.8 111.6 105.7 104.3 


Des Moines, Iowa 




Co. A 


100 79.4 57.3 48.9 46.6 


Co. B 


100 148.8 155.4 146.2 160.5 


San Francisco, Ca 


I. 


Co. A 


100 80.8 69.9 46.6 52.1 


Co. B 


100 121.8 125.4 127.4 118.3 


Salt Lake City, Utah 


Co. A 


100 97.1 77.1 77.1 54.3 


Co. B 


100 109.6 96.8 107.2 100.8 



Thus, if one is to account for the shifts 
in the national market share of a company, 
he must account for declines in some mar- 
kets and increases in others. To put it 
mildly, this is a messy (though apparent- 
ly inescapable) way of explaining why a 
firm's national market share changed. 

There are some important implications 
to be drawn from the wide variations in 
any firm's market shares from locality to 
locality, and from the diverse changes in 
shares of individual markets from one year 
to the next: 

(1) These variations might provide a 
basis for measuring the contribution of the 
''home office" to the changing fortunes of 
the company — in contrast with the things 
that "happen in the field." Indeed, the 
distinction between "things done by 
the home office" and "things done in the 
field" might be valuable in itself for diag- 
nosing difficulties and pinpointing suc- 
cesses. 

(2) It is pointless and misleading to 
offer explanations of shifts in national mar- 
ket shares since they do not parallel what 
happens in almost any individual market. 
National market share changes represent 
averages, often not very descriptive at that, 
of what happens in individual markets. 
Consequently, one must explain shifts in 
individual markets to account for move- 
ments in the average. 

( 3 ) By accounting for the market share 
shifts in local markets, it is possible to ex- 
pand enormously the number of "cases" or 
"observations" that one has for study. As a 
result, an analysis of variations in individ- 
ual market share changes may provide a 
basis for sorting out the effects of individ- 
ual sales promotion efforts employed by 
producers and by individual distributors. 

August 9. I think I have now arrived at 
some further implications that can be 
drawn from my work to date, which 
should be of special interest to business 
executives: 



THE NATURE OF MARKETING 



37 



( 1 ) Variations in market penetration at 
any given time and dissimilar changes in 
local market shares from year to year sug- 
gest that marketing executives should treat 
each market separately, wherever possible 
— rather than adopt common policies for 
the entire nation. Some executives, by 
thinking of their companies in terms of 
national market shares, take a wholly un- 
realistic view of their situation. Their 
policies should be related to the segment 
of the market they actually do serve, 
rather than being related to the national 
average. 

(2) I strongly conclude at this point 
that distributor strength represents the 
main source of a major appliance manu- 
facturer's market power. That is, it does 
once he has established both satisfactory 
set performance and acceptance of his 
product as a national brand whose price is 
in line with rival sets. Of course, this con- 
clusion raises but does not answer the cru- 
cial question of how a manufacturer 
achieves distributor strength. 

It is now clear that instead of selecting 
about 10 cases of substantial shifts in na- 
tional market share for explanation, I must 
revise the project to revolve about individ- 
ual market experience. But what form 
should the study take? 



STAGE 5 

Autumn 1955 through Summer J 956. A 
period of sweet promise when it appears 
that journalistic methods can be replaced 
by rigorous statistical techniques. 

October 27. Instead of considering the 
shifts in national market shares of approxi- 
mately 10 manufacturers as the universe of 
experiments to be explained, I now can 
work with shifts in local market shares of 
the same manufacturers in over 100 com- 
munities. In addition, I now have the op- 
portunity to make cross-sectional analyses 
which might enable me to explain differ- 



ences at any one time in the level of mar- 
ket penetration of any and all manufac- 
turers. 

With the enormous increase in the 
number of "observations" on which I can 
base my conclusions, the entire project has 
assumed an altogether different aspect. It 
will now be possible to have the "law of 
large numbers" working on my side. With 
information from over 100 markets, ran- 
dom and exceptional developments pre- 
sumably will cancel out and reveal the 
main effect of the sales actions taken by 
individual manufacturers. 

The method of approach that seems 
most suitable for handling these data is 
"multiple correlation." To use it, one 
must convert the "explanatory variables" 
(possible causative factors) into numerical 
values. Despite the difficulties encountered 
in Stage 2 in obtaining accurate measures 
of prices, product quality, dealer-distribu- 
tor margins, and the like, I hope it is pos- 
sible to obtain data that are sufficiently 
reliable to be used for this purpose. 

Despite the disappointments of finding 
no careful studies by individual manufac- 
turers of their decline in market position 
and the impossibility of explaining na- 
tional market share changes without look- 
ing at individual markets, I am pleased 
with recent developments. Now the in- 
vestigation is on a sound empirical foot- 
ing, and it should produce duplicable 
conclusions. That is, other researchers 
should reach essentially the same conclu- 
sions and obtain the identical results if 
they use the methods I am using. 4 In 
short, I may have come to a "scientific" 
method of isolating the factors accounting 
for sales success and for their relative im- 
portance. 

January 5. There are certain questions 
that puzzle me now. Is there a relation- 



4 Such duplicability, or, as it is often called, 
replicability, is considered to be one of the 
hallmarks of scientific methodology. 



38 



THE NATURE OF MARKETING 



ship between this study and "scientific 
marketing" in general? If so, what is the 
antithesis to scientific marketing? Does my 
project represent a test of the power and 
applicability of the scientific method for 
marketing executives? 

It would seem that one could not claim 
much for the scientific method in market- 
ing if, on hindsight, it would not be able 
to explain what actions succeeded and 
what ones failed in a specific industry. 

On the other hand, if one could de- 
termine what caused major changes in 
market shares, he might demonstrate the 
superiority of the scientific method over 
business experience and journalistic meth- 
ods — both working on the basis of hind- 
sight. And what one learns from hindsight 
provides much of the basis on which one 
attempts to exert foresight; any method 
that explains the past should help to fore- 
cast the future. 

As the study begins to assume a fairly 
pure statistical form, it becomes more 
difficult to understand why businessmen 
do not use rigorous quantitative methods 
much more than they do to arrive at their 
decisions rather than relying on intuition 
and "business judgment." 

July 21. The time has now come for me 
to formulate the statistical analysis pre- 
cisely so that I can grind out an answer. 
The problem has a familiar cast: the de- 
pendent variable will be the market shares 
of those TV set manufacturers for which 
I have information. Each market and each 
manufacturer will be treated as the phe- 
nomena to be explained. The independent 
variables will be such things as price, prod- 
uct quality, expenditures on advertising, 
number of retailers, and number of dis- 
tributor salesmen for each manufacturer. 

The conclusion will take the form of an 
estimating equation from which one can 
deduce the average change in market share 
associated with each $1,000 outlay of ad- 
vertising, each added retailer, and so forth. 



One can also derive coefficients of partial 
correlation which measure the relative im- 
portance of each independent variable, 
and likewise can compute standard errors 
which measure the reliability of the esti- 
mating equation. 

What specific form should the depend- 
ent variable — the phenomenon to be ex- 
plained — take? Should the market shares 
of many manufacturers — or only one — be 
used? Underlying this question are several 
problems: 

(1) Are the determinants of unit sales 
the same for all companies? Do variations 
in price, advertising, number of retailers, 
and so on, have the same effect on the 
market shares of all firms in an industry — 
even in an individual market? Or are the 
sales of some firms particularly sensitive to 
price incentives, while the sales of others 
respond most directly to variations in sales 
promotion efforts, and the sales of still 
others respond to changes in product per- 
formance features? 

(2) Is it appropriate to use "absolute" 
market shares as the dependent variable 
(assuming one is prepared to ignore the 
fact that market shares do not necessarily 
parallel short- or long-term profitability), 
or should the change in market share from 
the preceding year be used? If the latter, 
then one presumably would use changes in 
price, in advertising expenditure, in prod- 
uct quality, and in the number of retailers. 

STAGE 6 

Fall 1956 through Spring J 957. The sta- 
tistical technique is refined and prepara- 
tions made to grind out an answer. 

October 12. Marketing men (both busi- 
ness practitioners and academicians) differ 
from price theorists in that they draw 
different conclusions about the similarity 
of demand elasticities for rival firms that 
sell different brands of the same product. 

On the one hand, price theorists essen- 
tially ignore the problem and thereby tend 



THE NATURE OF MARKETING 



39 



to imply similar, if not equal, elasticities 
of demand for different producers of the 
same item. 

Marketing specialists, on the other 
hand, emphasize that businesses do and 
should stress different devices to win 
patronage. That is, their "marketing 
mixes" — the bundle of sales promotion 
devices like price, advertising, product 
quality, resellers' margins, personal selling, 
and credit accommodation — stress differ- 
ent appeals for patronage which con- 
sciously cultivate the patronage of differ- 
ent types of customers. Furthermore, 
some businessmen consciously feature 
sales promotion devices that will gain 
them the business of the types of custo- 
mers their rivals relatively ignore. 

I find it difficult to refute the position 
taken by the marketing specialist. Yet it 
leads inescapably to a most unhappy con- 
clusion. I can use the market shares of 
only a single firm as the dependent vari- 
able in my statistical attempts to explain 
changes in market penetration. However, 
I could explain changes in the market 
shares of every TV set producer for which 
I had supporting information, by means 
of a separate study for each one. A com- 
parison of the results for different com- 
panies would be quite instructive in itself 
— though by now it is clear that few peo- 
ple of my generation will ever live long 
enough to get the information needed to 
support such a comparison. 

As to whether to use the absolute mar- 
ket share or its change from the preceding 
year, the choice is relatively easy again. 
The question really boils down to whether 
the market penetration of a firm in any 
locality depends on what it itself is doing 
at that time or whether strong external 
forces keep the firm from experiencing 
dramatic jumps or sharp falls in its sales 
penetration. In other words, does market 
inertia operate to dampen the effects of 
any sales effort? Contrariwise, do market 



forces develop momentum that can carry 
forward over a considerable period? 

It is dangerous to decide such an issue 
on the basis of personal impression or 
logic; it is a factual question and should 
be settled solely by reference to the facts. 
But since these facts are difficult to obtain, 
I would be wise to formulate the statistical 
analysis so that it would produce an accu- 
rate answer — regardless of whether or not 
market inertia is strong. By using changes 
in market share rather than absolute mar- 
ket share, I would risk very little. Factors 
that cannot explain changes in market 
share can hardly account for the absolute 
level of market shares, which, after all, 
come to be what they are by a series of 
changes. 

December 1. I still think I have settled 
to my satisfaction those problems related 
to the selection of the dependent variable. 
But what problems do I face in the selec- 
tion of the independent variables to be 
used? 

There are many ways of expressing the 
independent variables — price, advertising 
expenditure, and so forth. They can be ex- 
pressed as an absolute quantity, but this 
has already been ruled out. They can be 
expressed either as absolute changes from 
the preceding year or as percentage 
changes from the previous year. It is pos- 
sible, too, to express them as a ratio of the 
effort made by the firm under study to the 
average effort made by its competitors. 
(For example, one could use its share of 
total advertising effort relative to average 
advertising effort, and its price relative to 
the average price.) 

As a final alternative, one could express 
the independent variables as changes in 
the ratio of the firm's effort to the efforts 
of its competitors. (Thus, a firm that 
made 5 per cent of total industry adver- 
tising outlays in the first year and 6 per 
cent in the second would be expressed as 
showing an advertising effort of 1.20.) 






40 



THE NATURE OF MARKETING 



The results that one reaches about the 
importance of any factor in sales success 
would depend heavily on the way he ex- 
pressed the independent variables. Which 
of the alternatives is most defensible? 

Clearly, the effects of any firm's efforts 
to win customers cannot be assessed in a 
vacuum, as if the company were playing 
against "par," and not against competitors. 
The outcome of sales efforts obviously de- 
pends on how they compare in strength 
to the measures taken by competitors. 
Since the dependent variable is being ex- 
pressed as changes in market shares, it 
seems appropriate to use changes in the 
ratio of a firm's efforts to the efforts of 
the industry as a whole. I realize, though, 
that a pretty good case could be made for 
using the proportion of the industry's total 
effort. 

April 7. The research plan now seems 
complete, except for the gathering of in- 
formation needed to carry it through. Mul- 
tiple correlation will be employed to ac- 
count for the major changes in local 
market penetration by individual com- 
panies. That is, when a substantial shift 
has occurred in a firm's national market 
share, a close study will be made of many 
local markets in which it sells. The data 
for that company, reflecting both its shifts 
in market share from the preceding year 
plus any changes in the independent 
variables which might account for those 
shifts, will be combined in a cross-sectional 
analysis. (Cross-sectional studies analyze 
what happens in different places or to dif- 
ferent groups during one fairly narrow 
period of time.) 

This formulation of the project seems 
quite rigorous, but so did earlier formula- 
tions. Some bugs will certainly turn up in 
this one. Before pushing ahead along this 
front, I should list the lessons learned dur- 
ing this and the preceding stage: 

(1) From the standpoint of both re- 
searchers and businessmen, by far the most 



significant lesson learned in Stages 5 and 
6 is the dissimilarity of different firms in 
their responsiveness to changes in price, 
advertising, personal selling, and product 
quality changes. One cannot simply apply 
the outcome of actions taken by one firm 
in order to forecast the result of similar 
actions taken by another. And one cannot, 
therefore, compute multiple correlations 
using the market shares of many firms as 
the dependent variable. 

(2) The effects of any company's sales 
promotion device can be considered as a 
combination of forces that help to over- 
come present and future influences which 
serve to build customer reluctance to pur- 
chasing that company's product. That is, 
sales efforts generally must overcome the 
momentum generated by past actions of 
other firms; in so doing, they will prob- 
ably develop momentum that will exert 
some future effects on the market. 

Consequently, both a researcher and an 
executive must estimate the timing of the 
results produced by any sales effort. A sales 
program is sometimes justifiable simply be- 
cause it reverses or halts forces generated 
in favor of a rival; on the other hand, 
some pay off mainly because of their fa- 
vorable effects months and even years 
after the effort was made. But as feeble 
as even the fanciest statistical devices are 
for measuring these different effects, the 
impressionistic method employed by the 
executive is even more feeble. 

(3) The effects of any firm's sales ac- 
tivities depend quite as much on what its 
rivals are doing as on its own actions. 
Rivals ordinarily have it in their power to 
largely offset the firm's actions simply by 
emulating its behavior. 

STAGE 7 

Summer and Fall J 957; Winter J 958. 
Frustration and discovery, mostly frustra- 
tion. 



THE NATURE OF MARKETING 



41 



June 20. Despite all of the offers and 
promises of cooperation I received from 
members of the industry, very little data 
have been forthcoming. Although I have 
been lulled into optimism by their friendli- 
ness and promises of help, I must recog- 
nize that I will not get the information I 
require from the major companies in the 
industry. Also, it is quite clear that they 
do not even have some of the information 
that I require. 

For example, they do not know the 
prices charged by different retailers of their 
own sets in any market at any one time — 
let alone have a continuous picture of 
changes in retail prices. Naturally, they 
know even less about the prices charged 
by competitors. Moreover, their informa- 
tion about advertising activity by distribu- 
tors and dealers in local markets is frag- 
mentary and unreliable. As I found years 
ago, they barely have a clue to the margins 
obtained by distributors and dealers in 
individual markets. Conclusion? The proj- 
ect as I formulated it in the spring may 
have been rigorous and esthetically satisfy- 
ing, but to get the data it requires costs 
vastly more than I can afford. 

August 11, Now that I've had time to 
reflect on it, the way I've formulated the 
problem itself is deficient on at least two 
important counts: 

(1) It assumes that the effects of a 
firm's sales promotion efforts are the same 
in all markets in which it does business. 
This assumption is clearly not realistic for 
the TV set industry — or for the major 
appliance industry generally. Certain mar- 
kets (especially the large metropolitan 
ones) are extremely sensitive to price 
changes; others seem to be particularly 
sensitive to advertising. Similarly, market 
shares in small cities and towns seem to 
be determined more by the number and 
quality of dealers than by price and ad- 
vertising — according to prevailing opinion 
among sales executives in the industry. 



One conclusion suggested by these con- 
siderations is that each firm's market pene- 
tration must be explained separately in 
every market. A less congenial conclusion 
can hardly be imagined! 

It has been clear from the start that the 
essential character of the television set 
business has changed greatly over its rela- 
tively short life. As a result, one cannot 
regard observations drawn from different 
stages of the industry's history as repre- 
senting the same "universe." The forces at 
work in, say, 1952 are quite different from 
those operating now in 1957. Certainly 
the relative impact of such factors as price 
and advertising would probably be quite 
different in each of those two years. Con- 
sequently, one cannot safely use data for 
many successive years to derive measures 
of the influence of selected factors by 
formal statistical methods. Thus, the con- 
clusion that each market must be consid- 
ered a separate universe, especially when 
combined with the rapidly changing char- 
acter of the industry, appears to rule out 
the use of statistical methods. What seems 
called for is a retreat to journalism — if not 
to nihilism! 

(2) The other deficiency in the formu- 
lation decided on last spring is quite as 
grievous. It concerns the number and na- 
ture of the independent variables used to 
account for changes in market shares. 
From the start, I simply adopted the view- 
point of the economist and marketing spe- 
cialist and assumed that market penetra- 
tion was the result of "sales efforts" (price, 
advertising, personal selling, and so forth ) . 
Now two additional sets of forces seem 
almost equally likely to affect any firm's 
penetration of a local market: (a) the re- 
sources that both the distributor and the 
manufacturer devote to that market; (b) 
the managerial skills of the local distribu- 
tor and the manufacturer bearing on the 
various factors that affect sales success 
(that is, costs of operation, personal con- 






42 



THE NATURE OF MARKETING 



tacts, financial liquidity, access to credit, 
and the like). 

Business lore is full of stories about bril- 
liant men whose wise decisions came to 
naught because they were executed poorly. 
We all know stories about the ingenious 
marketer who launches an effective adver- 
tising program at the consumer precisely 
at the time when objectionable behavior 
on the part of the manufacturer's repre- 
sentatives has created a desire on the part 
of the distributors to "make the promo- 
tion lay an egg." 

The ability of a businessman to create 
a harmonious organization, and to develop 
the requisite skills among his employees 
and distributors, gives a greater impact to 
whatever the firm tries to do. Conse- 
quently, the managerial skills of the manu- 
facturer must figure significantly in any 
firm's market success or lack of it. This 
conclusion adds mightily to the already 
overwhelming disappointments met in this 
stage. 

March 17. I am rapidly coming to the 
conclusion that it is "every firm and every 
market and every year for itself." If each 
event is an isolated phenomenon, one 
simply cannot employ statistical tech- 
niques which, above all, assume an essen- 
tial similarity or regular and predictable 
change in underlying conditions. Neither 
of these conditions is met by activities in 
the television set industry. 

If the use of empirical methods to ac- 
count for sales success leads to such a 
dead end, what happens to executives who 
try to make their way with common sense 
and "business judgment" alone? Specifi- 
cally, on what basis do businessmen in the 
TV set industry decide which measures to 
adopt and which to reject in pursuit of 
greater sales? How do they — or don't they 
— overcome the pitfalls that have bedev- 
iled my efforts? 

One might try to answer these questions 
by taking selected decisions by executives 



in TV set manufacturing companies and 
studying them in great detail in order to 
determine on what basis they were made. 
Several studies of this sort have been 
made, I know, but they ordinarily deal 
with such relatively simple decisions as 
whether or not to buy an electronic com- 
puter. But even these apparently simple 
decisions turned out to be enormously 
complicated, and the test results are not 
duplicable. 

I'm afraid I don't have the stomach for 
such an undertaking, nor am I confident 
that it would be worthwhile, even if it 
were carried out with skill and vigor. In 
addition, it is useless to investigate the 
basis for executives' decisions, because the 
very nature of their method is mysterious. 
Business judgment must be defined as 
an inexplicit and at least partly noncon- 
scious method of reaching conclusions. 

Another approach would be to investi- 
gate whether executives are aware of the 
kinds of things which I have "discovered" 
in the course of the study. I refer to such 
things as the different elasticities of de- 
mand with respect to price and advertising 
for individual firms, variations in sources 
of success from market to market, and 
changes in the effectiveness of various sales 
promotion devices as the industry has ma- 
tured. Whether they understood the im- 
plications of these "discoveries" should 
also be investigated. Of course, it is very 
difficult to find out what a person does 
not understand. For whatever it is worth, 
my impression is that most of the execu- 
tives I met in the TV industry are not 
aware of these things and do not under- 
stand their relevance to their own business 
decisions. 

This conclusion raises a paradox that 
bothers some people. Many businessmen 
seem to be unaware of some quite crucial 
business facts and concepts. Nevertheless, 
they are generally richer than most people 
who are familiar with them. Also, they 



THE NATURE OF MARKETING 



43 



appear to discharge their jobs with what 
their associates consider real distinction— 
if thev didn't, would they not fail? So this 
superficial argument goes. 

But wait a minute. Just a little thought 
will show that the payoff in business de- 
pends on many things other than absolute 
skill. Even if luck is discounted, it is clear 
that relative ability rather than absolute 
ability matters most. Consequently, it is 
impossible to state that, since an execu- 
tive is successful, he understands "why 
what he does makes him successful." (A 
close connection between success and un- 
derstanding may exist when many execu- 
tives trained in the scientific approach to 
business hold positions of influence.) 

Moreover, it would be highly unlikely 
that the unconscious processes of execu- 
tives do not fall prey to some of the erro- 
neous assumptions that I have made at 
various stages in my research. For example, 
don't they "subconsciously" apply the les- 
sons learned from the experience of other 
firms to their own situations — even when 
thev do not apply? And do they not often 
make decisions in different markets with- 
out taking account of the differences 
among them? And, above all, are not their 
conclusions concocted from a stew com- 
posed of facts, impressions, assumptions, 
hearsay, and wishful thinking — completely 
unadjusted to take account of the identity 
of the firms involved, the stage of the 
industry, variations among local markets, 
and differences between what was decided 
and what was actually carried out? With 
due respect for the great superiority of the 
human brain over the electronic computer, 
it cannot produce brilliant conceptions 
and conclusions out of trash. 

There is small consolation, however, in 
finding that business judgment is even 
more fallible than the scientific method is. 
There must be, there must be, an escape 
from the nihilism of this Stage 7 via the 
scientific route. 



STAGE 8 

Summer 1958 to Summer J 959. Other 
methods of approach are formulated. 

July 28. Several avenues of escape now 
suggest themselves; none, however, pos- 
sesses the classic beauty of the multiple 
correlation models. The most promising 
approach probably would be to divide local 
markets into roughly homogeneous types 
— according to the size of the market, the 
number of rival distributors, the severity 
and form of competition, and the turn- 
over among dealers. One might divide 
local markets into groups simply by ask- 
ing major sales executives in the industry 
to classify them (those markets about 
which there was disagreement could be 
omitted altogether — for it is not necessary 
to use every single market). 

Once such market groupings are ob- 
tained, it should become possible to make 
cross-sectional studies employing multiple 
correlations based on the assumption that 
all the markets within each group are 
essentially similar. These studies would be 
formulated so that the independent vari- 
ables would be the changes in market 
share of one specific firm from the pre- 
ceding year in each market. These inde- 
pendent variables would be expressed as 
changes in the ratio of that firm's effort 
(price, advertising, retailing representation, 
and so forth ) to those made by all of its 
rivals. Among the independent variables 
used to explain shifts in market share 
would be resources possessed by distribu- 
tors and the managerial skill of manufac- 
turers and distributors. (These last two 
presumably would be expressed in quanti- 
tative form by the use of scaling tech- 
niques; that is, the firm's resources and 
managerial skills might be rated from A to 
E, assuming that C represented the aver- 
age for all firms in that market.) 

The use of such a method should yield 



44 



THE NATURE OF MARKETING 



separate hypotheses (or estimating equa- 
tions) for each firm in each class of mar- 
ket. It should also be possible to compute 
such equations for many different firms in 
the same class of market and to observe 
the differences in determinants of success. 
Hopefully, the results would suggest hy- 
potheses to account for the differences 
that emerged. 

A much more ambitious solution would 
be to experiment with simulations of re- 
ality by means of an electronic computer. 
One might formulate hypotheses about 
the relative influence and nature of the 
forces affecting market shares in particular 
types of markets and test them on the 
machine. By trial and error, one might 
find a set of hypotheses (in equation 
form) that would explain satisfactorily 
what actually happened in each class of 
market. These hypotheses would repre- 
sent tentative solutions, to be matched 
against future experience for verification. 

April 11. I have grown somewhat sour 
on both of these approaches. Moreover, 
they would require vastly more informa- 



tion than I have been able to obtain thus 
far. To employ either of them would un- 
questionably require special arrangements 
being made to collect in advance the in- 
formation that is needed. At least a two- 
year period would be required, I am sure. 
Besides, the financial and time costs of 
such an undertaking are far greater than I 
can afford. 

So, whether I like it or not, this prob- 
ably represents the end of the line for me 
on this research project. 5 



Does this approach contradict or re- 
inforce the approach described in 1-1? 
Was this researcher really scientific? What 
further could he have done, either of a 
scientific or a practical nature, to realize 
his objective? 



5 Professor Oxenfeldt, in the full report, 
uses his experience in this project to de- 
velop a set of generalized conclusions 
about decision making and marketing re- 
search. 



THE NATURE OF MARKETING 



45 



Harry Higbee, salesman for the Win- 
way Hat Company, is talking to Bill Burn- 
ham, buyer for hats in the Boling Bros. 
Department Store, catering to the me- 
dium-to-high income group. 

Salesman: Mr. Burnham, I appreciate the 
opportunity to talk to you 
about carrying our Winway 
hats. This is the kind of store 
we'd be proud to have them 
in. 

Buyer: That's interesting. There was 
a man in here several weeks 
ago who was asking if we 
didn't have a Winway hat in 
his size. But he took another 
brand, without much fuss. 

Salesman: How many brands do you carry 
now? 

Buyer: We have the Londoner and 
the Balmak — both good brands. 
Our customers like them. 

Salesman: They are good brands. But 
why don't you round out your 
line with the Winway? Our 
price range is lower, distinctly 
lower than either of them, too. 
Then you'd be able to satisfy 
all your customers, and maybe 
attract some new ones. 

Buyer: That means increasing inven- 
tory about 50 per cent, and I 
doubt if we would get enough 
extra sales to justify it. I'm 
sort of cramped for space, too, 
as you can see. 

Salesman: We had an independent mar- 
keting research agency make 



CASE 1 -3 

Decision at Boling Bros. 
Department Store 

a survey for us among men hat 
buyers. Their brand prefer- 
ences came out like this — see 
on this chart — Winway 25 per 
cent, Balmak 20 per cent, Lon- 
doner only 5 per cent — I guess 
they think that's a foreign 
make — and others, all together, 
10 per cent. 

Buyer: What about the rest — let's see, 
25 and 20 and 5 and 10 is 60 
— the remaining 40 per cent? 

Salesman: They're classified as "don't 
knows." Couldn't make up 
their minds, I guess. 

Buyer: Well, all I know is those fig- 
ures don't apply here. The 
Londoner gets 70 per cent of 
our business, and the Balmak 
30 per cent. Also, this is the 
case even though, as I'm sure 
you know, the Londoner runs 
a few cents higher than the 
Balmak. 

Salesman: That's because you don't carry 
the Winway. Why, in the 
Tuscary Department Store — 
only 50 miles away — the Win- 
way is enjoying 80 per cent 
of the business. You've seen 
our advertisements in Esquire, 
haven't you? 

Buyer: Oh, yes. Nice ads. But our 
trade is accustomed to the two 
we have. . . . 

Salesman: Well, keep us in mind. I'll 
stop in to see you my next trip 
through here. And in the 
meantime, why don't you ask 



46 THE NATURE OF MARKETING 

some of your customers what should he apply? From the facts revealed 
they think of Winway, and see in the course of the above conversation, 
what they say? and your own observations of men's hat 

buying, what do you think determines the 

kind of hat the average man buys and, if 

Should Burnham take on the Winway different, what determines the kind of hat 

hat? What kind of considerations should the Boling Bros, customer buys? What do 

affect his decision? What kind of analysis you think of Higbee's salesmanship? 



BIBLIOGRAPHY TO SECTION ONE 

Alderson, Wroe, Marketing Behavior and Executive Action. Homewood, 111.: 
Richard D. Irwin, Inc., 1957. 

Britt, Steuart Henderson, The Spenders (in rebuttal of Vance Packard; see be- 
low). New York: McGraw-Hill Book Company, Inc., 1960. 

Bursk, Edward C, Donald T. Clark, and Ralph W. Hidy, editors, World of 
Business, Volume I (a collection of historical and literary writings about 
marketing from 2000 b.c. to the present). New York: Simon and Schuster, 
Inc., 1962. 

The Editors of Fortune, Markets of the Sixties. New York: Harper & Brothers, 
1960. 

Mazur, Paul, The Standards We Raise: Dynamics of Consumption. New York: 
Harper & Brothers, 1953. 

McNair, Malcolm P., and Harry L. Hansen, Readings in Marketing. New York: 
McGraw-Hill Book Company, Inc., 1956. 

Packard, Vance, Hidden Persuaders (attack on advertising practices). New York: 
David McKay Company, Inc., 1957. 

Potter, David M., People of Plenty (a sociologist's analysis of "America's institu- 
tion of abundance," advertising). Chicago: University of Chicago Press, 1954. 

Schlaifer, Robert, Probability and Statistics for Business Decisions. New York: Mc- 
Graw-Hill Book Company, Inc., 1959. 

The Twentieth Century Fund, Does Distribution Cost Too Much? New York: 
The Twentieth Century Fund, 1939. 

Vaile, Roland S., Ewald T. Grether, and Reavis Cox, Marketing in the American 
Economy. New York: The Ronald Press Company, 1952. 

Wright, David McCord, Capitalism. New York: McGraw-Hill Book Company, 
Inc., 1951. 



SECTION TWO 



Understanding 
the Consumer 



As we proceed, we shall be confronted with the complexity of all the different 
courses of action that marketers have to choose from and mix together into a 
competitive strategy; and with the complexity that stems from competition itself, 
in which action occasions counteraction in an endless chain of offensive and de- 
fensive moves. 

As part of this first stage of our exploration, however, we are concerned with 
the complexity of the raw materials the marketer must work with— that is, people, 
meaning both the ultimate consumer and all the intervening salesmen and middle- 
men (who likewise can have a decisive effect on sales by their interest and en- 
thusiasm, or lack thereof). 

In this section and the next we shall consider (a) the forces affecting the 
buying behavior of the countless individuals who by their cumulative, changing, 
and sometimes conflicting attitudes make up the phenomenon of a market; and 
(b) the difficulty and desirability of getting information about them, so as to 
have a basis, in turn, for planning marketing action to affect their attitudes. 

NATURE OF DEMAND 

Let us take people first in the aggregate: how many will buy under different 
circumstances? 

The backbone of marketing is demand. This is not demand in the usual sense 
of people insisting on something. Rather, it is the economists' term for people's 
ability and desire to buy goods and services, as contrasted with the supply of goods 
and services that are available. However— and here is where marketing adds its dy- 
namic dimension of change— demand is not something that the manufacturer or 

47 



48 UNDERSTANDING THE CONSUMER 

supplier must accept as it is. Marketing causes things to happen; by marketing, a 
business takes action so that when the demand for its goods and services is non- 
existent or so passive as not to deserve the name, then demand can be generated; 
and when the demand is of any given strength, then it can be made stronger. 

One of the things that makes marketing so complex is that the demand for a 
product or service depends on a whole range of factors. Some of these factors 
are out of the direct control of the marketer, and some of them he can attack 
and change as a way of building or increasing sales. 

But even in the case of those that are not directly subject to his influence, 
the marketer can shape his action for the sake of sales. For example, there is 
little the marketer can do to change the number of people who live within the 
range of his influence, or little he can do about the size of their incomes or their 
physical capacity to consume. Thus, suppose there are x number of people in his 
market, and he is one of 100 sellers of golf clubs, each of whom is enjoying one 
per cent of the market. He cannot increase the number of people who are there to 
sell to. But suppose ten per cent of the golfers play left-handed. If he offers a 
left-handed golf club for sale (and his competitors don't), he has a chance of get- 
ting not one per cent but ten per cent of the market— just by fitting his product 
to an existing market. 

On the other hand, he is now definitely limited to ten per cent of the mar- 
ket; or maybe, in real life, it is an even smaller and leaner group of prospects. 
His alternative is to go after sales, not by making his product fit the existing 
market, but by changing the market itself: not the number of people or their 
characteristics (which he can't change) but their wants and preferences for golf 
clubs— that is, induce them to prefer his brand to the others' brands. 

To take another example, suppose he is selling tickets to a show. Within trav- 
eling distance of the theater are 100 people who can afford to pay $5.00, and 
1,000 people who can afford $1.00. He cannot change the amount of money in 
their wallets. But he can shape his actions to the circumstances; he can figure 
that if he prices his tickets at $5.00, the most he can collect is $500—100 at $5.00 
—while at $1.00 he can sell 1,000 for a total of $1,000 or twice as much. (This is 
what economists call elasticity of demand. That is, so many more people will 
buy at a lower price that the seller actually takes in more total money at the 
lower price; or, conversely, so many fewer people will buy at a higher price that he 
takes in less total money at the higher price.) 

Here again we get into the possibilities of not merely taking existing physical 
circumstances as they are, but trying to change some of the psychological factors. 
Suppose an additional 500 people, farther away from the theater, would have to 
spend another $1.00 in transportation; they can afford to, but they don't feel like 
it. By advertising, by making his show sound that much more exciting and de- 
sirable, he changes those people's desire, that is, increases their desire, to the point 
where some of them pay not just $1.00 for the ticket but also $1.00 for transporta- 
tion. (This then becomes an example, not of taking advantage of the existing, 



UNDERSTANDING THE CONSUMER 49 

and as we have seen, elastic demand, but of actually expanding demand. A good 
part of marketing effort in this country is devoted to this kind of action— ex- 
panding, increasing, people's desires and wants for the product or the brand that 
the marketer is selling.) 

And it is in this area of the psychological— wants, desires, attitudes— that we 
must cope with (or, from the point of view of a successful marketer, capitalize on) 
many factors which are less tangible, less readily discernible, and which therefore 
demand more subtle understanding and yield only to more scientific analysis. 



CONSUMER BEHAVIOR 

How can we focus on people as the various and variable individuals that they 
are? Let us begin by trying to understand why people buy or do not buy a par- 
ticular product. 

Many managements, who think they know, are actually in the dark as to how 
their products are regarded by present and potential customers. A classic example 
is what happened to sellers of French-type champagne when they embarked on a 
program to build a market in this country. Ask people what they like— a sweet or 
a dry champagne. They will invariably answer "dry," because that sounds like the 
more sophisticated answer. The wine people did ask, did get that answer, and 
took consumers at their word; sales did not develop. The trouble is that Americans 
really like sweet drinks— whether they themselves know it or not. Research finally 
revealed this fact. Champagne has now become very popular in this country — 
sweet champagne, that is. 

Despite the slipperiness of facts about how people may regard a given 
product or brand, many managements go ahead making "seat-of-the-pants" deci- 
sions on product design, packaging, pricing, and promotion which entail large 
sums of money and, in some cases, determine the success or failure of the whole 
enterprise. 

A certain amount of hit-or-miss, trial-and-error tactics is inevitable. Consumers 
are individual human beings. They are all different from one another; and even 
the same people vary in their feelings from time to time. One family may want 
a car for show, or social prestige; another for economical transportation. Within 
the same family, the father may think of the car's mechanical features, the mother 
of its comfort, and the teen-agers of its dating possibilities. And even the same 
individual— say, the father— may have one attitude toward the car for trips, and 
another for commuting to work. Note in Table 2-A the different weight of factors 
influencing the purchase of boxed chocolates when bought for different purposes 
and by different age groups. For example, buyers consider the package less im- 
portant than the assortment, brand name, etc., but the package gains in im- 
portance when the candy is bought as a gift, though not so much in the case of 
buyers over 50 years old. 






50 UNDERSTANDING THE CONSUMER 



Psychological vs. Logical 



People buy things for many reasons— some tangible, some intangible; some 
conscious, some unconscious. To the marketer, these needs or wants, these motives 



TABLE 2- A. CONSUMERS' RATING OF FACTORS THEY CONSIDER IMPORTANT WHEN 
BUYING BOXED CHOCOLATES FOR GIFT AND FOR OWN USE 

(as percentage of buyers' rating each factor) 



Assc 


rtment Brand Name Pa 
)wn Use Gift Own Use Gift C 

97% 88% 81% 54% 
92 84 77 59 
95 84 83 36 


ckage 


Store 
Owners ' 
Recommen- 
Price Quality dation 


Gift C 

Under 30 

years old 93% 
30-49 

years 93 
50 years 

and over 84 


)wn Use Gift 

28% 67% 

25 65 

26 61 


Own Use Gift 

80% 98% 
65 97 
65 92 


Own Use Gift Own Use 

93% 21% 18% 
95 18 17 
83 15 16 



are not always logical; they are not the motives that he would have (at least he 
does not think so, though he may be mistaken about himself too). That is why 
it is so important to remember that buyers have "psychological" rather than 
"logical" reasons for buying, or for being susceptible to a selling attempt. Or, to 
put it another way, they consider what they do is logical; their logic is their own, 
and it may happen to be different from the seller's. 

Thus, two authorities in psychology, Donald Snygg and Arthur W. Combs, 
in their book on Individual Behavior, make this observation: 

From the point of view of an observer who knows the location of an 
exit, the behavior of a fire victim in rushing back again and again to a jammed 
door is completely unreasonable. From the point of view of the victim in 
those circumstances, it is the most reasonable thing he can do because the 
door is the closest approximation to an exit that he can find. However ca- 
pricious, irrelevant, and irrational his behavior may appear to an outsider, 
from his point of view at that instant his behavior is purposeful, relevant, and 
pertinent to the situation as he understands it. What it looks like to others 
has no bearing upon the causes of his behavior. 1 

So one of the main attributes of good marketing is to be sure not to stop 
short with what seems to be the logical, surface explanation of consumer behavior. 
The trouble is that people, with all sorts of different and conflicting motives, do 

1 Donald Snygg and Arthur W. Combs, Individual Behavior (New York, Harper & Brothers, 
1949), pp. 12-13. 



UNDERSTANDING THE CONSUMER 51 

not always know themselves what their motives are; and, even if they do, they 
will not always reveal their thoughts under direct questioning. 

The fact that consumers are scattered geographically and are not under the 
direct control of the marketer simply adds to the problem. 

This buying for psychological reasons is true of both consumer goods and 
industrial goods, although, in the latter case, the buyer is likely to follow business 
reasoning more than personal attitudes. Nonetheless, he is subject to flattery, the 
need for self-assurance, and so on. A great many industrial salesmen overlook the 
possibilities along this line. 

How Emotional? 

We are considering here the question of why consumers behave as they do 
in buying or not buying. While it is perfectly clear that their attitudes often 
have an emotional base, the question of what to do about it is another matter. 
Although selling strategists tend strongly to favor emotional appeals, communica- 
tion experts do not altogether agree. According to the authoritative book, Com- 
munication and Persuasion, published by the Yale University Press, there are a 
number of experimental studies that cast doubt on the superiority of "emotional" 
appeals over rational appeals. 2 

Some of the difference of opinion may result from the fact that it is a mistake 
to generalize on either side of this question; neither emotional appeals nor ra- 
tional appeals are right in all cases. (We are discussing the content or theme of 
the appeal, not the way it is presented. Actually, a rational appeal can be pre- 
sented quite strikingly and colorfully, just as an emotional appeal can be presented 
very dully and flatly.) As we shall see in Section Four, where we tackle the prob- 
lem of stimulating buying behavior, a combination of emotional and rational ap- 
peals may be the most effective; the particular product-market situation will de- 
termine what is optimum— and it will vary the whole way from mostly emotional 
to mostly rational. 

The degree of emotionalism or rationality of the appeals will depend in part 
on the kind of emotion involved in the attitudes. For example: 

When people have anxiety (like a dread of pain or loss of teeth), to back 
a product (as in the dental-hygiene area) with a strong emotional fear cam- 
paign is likely to be less effective than a more rational one using a minimal 
amount of fear appeal and a large amount of neutral information. 3 Con- 
versely, when people have a comfortable emotion, like pride of home and 
family, a positive emotional appeal can be expected to have a favorable 
effect. 



2 Carl I. Hovland, Irving L. Janis, and Harold H. Kelley, Communication and Persuasion 
(New Haven, Yale University Press, 1953), p. 57. 

3 Irving L. Janis and Seymour Feshbach, "Effects of Fear-Arousing Communications," 
Journal of Abnormal and Social Psychology, Volume 48, 1953, pp. 78-92. 



52 UNDERSTANDING THE CONSUMER 

That people who are low in gregariousness or who are likely to be "blue" 
and depressed from time to time are more likely to be subject to appeals 
which are positively or even romantically sentimental, can be inferred from 
findings of Professor Elihu Katz and Paul F. Lazarsfeld, pioneers in the field 
of communication and social research at Columbia University. 4 On the other 
hand, people who are better educated are more likely to be influenced by 
rational appeals, everything else being equal. 

In other words, even though the marketer is not necessarily going to use 
out-and-out emotional appeals, he needs to understand the emotions that in- 
fluence people's buying behavior. 



Social Influences 

Not all of a consumer's motives are generated within himself. In one sense, 
none of them are, since even the individual's ability to think as an individual is 
conditioned by his environment and by his upbringing and education. More spe- 
cifically, a great many of his individual actions are attempts to adjust to the stand- 
ards of people around him, and to increase his feelings of pride and prestige in 
terms of the way he wants others to see him. 

Here, again, the buyer or intended buyer is not always conscious of why he 
does what he does. He may buy a large automobile because he really wants to be 
regarded as prosperous as others in his neighborhood, but he may fool himself 
into thinking that the reason he bought it was because he made such a good deal, 
or the upkeep will be less in the long run. There are some researchers who feel 
that these social influences are the most powerful of all motives; and, as long as we 
are thinking about people who live above the level of necessity, as most people 
do today, it is probable that they are more concerned with their relations with 
other people than with their bodily needs and satisfactions. 

According to Pierre Martineau, Director of Research and Marketing for the 
Chicago Tribune: 

There is certainly a rough correlation between income and social class. 
But social class is a much richer dimension of meaning. There are so many 
facets of behavior which are explicable only on a basis of social class dynamics. 
For instance, this analysis of the purchase of household appliances in Chicago 
over a four-year period shows a very different picture by income and by class. 
Income analysis shows that the lowest income group represents an un- 
derstandably smaller market, but nevertheless a market. Social-class analysis 
highlights a fundamental difference in attitudes toward the home between 
the two lower classes. The Upper-Lower Class man sees his home as his castle, 
his anchor to the world, and he loads it down with hardware— solid heavy 



4 Elihu Katz and Paul F. Lazarsfeld, Personal Influence (New York, The Free Press of 
Glencoe, 1955), p. 378. 



UNDERSTANDING THE CONSUMER 53 



NINE APPLIANCE TYPES- FOUR- YEAR 
PERIOD 



By Income 




Over $7,000 
4,000-6,999 
Under 4,000 


36.2% 

46.0% 

17.8% 

100 % 


By Social Class 




Upper and Upper- Middle 
Lower-Middle 
Upper- Lower 
Lower- Lower 


16.6% 

29.2% 

45.7% 

8.5% 



100 % 

appliances— as his symbols of security. The Lower-Lower Class individual is 
far less interested in his castle, and is more likely to spend his income for 
flashy clothes or an automobile. He is less property-minded, and he has less 
feeling about buying and maintaining a home. 5 

Fashion 

The very fact that consumers respond to social influences means that they are 
likely to act differently or to feel differently from year to year, or from season to 
season. Thus, small cars become more socially acceptable than they used to be, 
and so there is a trend toward greater purchases of them. 

Fashions— by which we mean not just fashions in women's styles but fashions 
in automobiles and cigarettes and- even machine tools, and so on— are often 
cumulative in their effect. A small number of people of high prestige may adopt 
a certain style to be "ahead of" the rest of the world, and then everybody else 
begins to adopt the same style to "catch up" with the others. 

Yet even the phenomenon of fashion— which appears on the surface a hodge- 
podge—is amenable to analytical thinking. Alfred H. Daniels, Vice President of 
the Federated Department Stores, Inc., has described a revealing incident that 
happened to him when he was Merchandising Vice President at Abraham & 
Straus in Brooklyn, New York: 

[One time] our Blouse Department got "very sick." It was serious be- 
cause we were going into our big Christmas season, during which a continued 
slump of the same degree would have been disastrous. Not being too close 
to the facts, I went at the problem administratively. I talked to the buyer 
and divisional merchandise manager, both of whom were rightfully dejected; 
called a few friends, who had no suggestions; had our competition shopped 



5 Pierre Martineau, "Social Classes and Spending Behavior," Journal of Marketing, October 
1958, pp. 125-126. 



54 UNDERSTANDING THE CONSUMER 

and found there was "no activity"; and had a complete survey made, finding 
that we were "covered." The sum total of my administrative efforts, there- 
fore, was nothing. 

As a last resort it should have been the first— I requested the unit mer- 
chandise cards, together with samples, of the best sellers. Upon scrutiny, it 
was apparent that seven were wool jerseys. In desperation, we decided to run 
a wool jersey promotion. This decision, though, did not quite "hit home." 
Competitors had advertised wool jerseys, and our grapevine had indicated 
nothing startling. As our meeting was about to adjourn, I asked to see the 
two best sellers once more. Then the light came. Never did two necklines 
plunge so precipitously— and wickedly! 

No time for modeling. We hurriedly reordered the two blouses in large 
amounts, bought three new styles with the same common denominator, 
and scooped— both literally and figuratively, I guess— the country with "Deep 
Plunging Wool Jersey Blouses." Our blouse business became enormous. Some 
even said that a new fashion had been started, for the "deep plunging" look 
was soon advertised in dresses, slips, and bras all over the country. 

Had I "started" a fashion? Of course not! There had been a general, and 
obvious, fashion trend toward bareness— "The Bare Look"— for some time. 
Subconsciously, that is probably why I had asked to see the two blouses again. 
Had I started a specific style? Of course not. We all had some blouses of this 
kind— and they had been selling. . . . 

It will be seen that much of what I had been discussing here is scientific, 
if by scientific one means a logical interpretation of merchandise facts. 6 
A further help to understanding and predicting fashion is that it has a certain 
rhythm. It tends to go from one extreme to another over a period of time. Here 
is a good example of a concept serving as a model, and facilitating systematic 
analysis. According to Professor Dwight E. Robinson of the University of Wash- 
ington, the behavioral complex underlying all stylistic innovation— that is, all 
changes in design which are not purely the results of engineering advances— is 
the pursuit of novelty for its own sake. 

Paul Poiret, the top Paris couturier of the 1920's, once summed up his credo 
by declaring, "all fashions end in excess"— a principle which is the beginning of 
wisdom for all who are concerned with style policy. He was aware that the over- 
riding responsibility of the designer in a fashion market is the unending pro- 
vision of novelty. Implicitly, he recognized that one of the most exacting problems 
the stylist ever faces is that of deciding what to do when he has exhausted the 
possibilities of a current direction in styling emphasis. What does he do, for ex- 
ample, when the waistline, hemline, or any other line has been carried as far as 
it will go? It is here that the couturier must exercise to the utmost every ounce of 
his insight into the meaning of fashion. 



6 Alfred H. Daniels, "Fashion Merchandising," Harvard Business Review, May-June 1951, 
p. 56. 



UNDERSTANDING THE CONSUMER 55 

As a couturier, Poiret knew that the appetite for novelty, arising from the 
twofold insistence of the lady of fashion on preserving her inimitability from the 
onslaught of the vulgar and on demonstrating her affluence through unrelenting 
expenditure on newly cut costumes, is never satisfied with any one mode of pre- 
senting the figure. Fashions in dress design subsist on measures to transform or to 
distort, whether through exaggeration or minimization, the shape and features of 
the human figure. To illustrate: 

The hoop skirts of the eighteenth century and the crinoline of the nine- 
teenth ballooned to diameters of eight feet. As the hip line was exaggerated 
beyond the point of simply imperiling navigation, to the point of making it 
literally impossible, the waistline was tightened to the point of suffocation 
and interference with digestion. 

In the interim, in the Directoire and Empire periods around 1800, a con- 
trary tendency toward undress was exploited. La Parisienne's test of the 
suitability of a pseudoclassical gown of transparent silk was to see whether 
Madame could easily draw it through a ring taken from her little finger. The 
last resort of modesty was flesh-colored tights. 

The flapper of the Jazz Age, though she was, in her turn, the despair 
of her late- Victorian parents, never dashed about in quite the dishabille of 
Mme. Recamier, but what she missed in transparency she made up for with 
leg display. 

Whether, at a given time, the particular form of emphasis is toward padding 
out or constricting, toward concealment or exposure, once such a movement is 
launched it must be intensified each season, ensuring that the ultrafashionable will 
be able to disport themselves in more of a good thing than their less-favored con- 
temporaries. Indeed, this recurrent pattern can be traced back to even earlier 
centuries. The style of European costume associated with the French Regency 
(1715-1730) emphasized delicacy and restraint. Its graceful, free-flowing costume 
is best remembered in the airy, idyllic scenes depicted in the paintings of Watteau. 
Yet it fell between the ornate periods of the stately Baroque and the frivolously 
extravagant Rococo. 

The most important corollary of Poiret's axiom is this: a fashion can never 
retreat gradually and in good order. Like a dictator, it must always expand its 
aggressions— or collapse. Old fashions never just fade away; they die suddenly and 
arbitrarily. The reason is simple and logically inescapable. The one thing fashion 
cannot stand is to repeat the recently outmoded style, the passe. Better for the 
lady of fashion to look like a freak than to be mistaken for her grocer's wife dolled 
up in a cheap version of something she herself sported a year or two ago. 

The rule that fashion never smoothly retraces its footsteps demonstrates itself 
even in the history of architecture, which is a particularly stern testing ground. 
Look at what happened to the arch when, at the end of the late Gothic style, 
it had reached the extreme of pointedness. Did the Tudor architects, seeing that 
matters had reached a geometric impasse, decide to blunt it just a bit? Not at all. 








X£ 



Ql 



SMt 




Tl"^ 




-r 




— ,-j 



r 




»f 



Nl 



UJ 






t> 



en 

z 
o 



HI 



UJ 



Hi 



.2*311 

l°s| 

Hjo 

loll 
|33g 

^"2-b v 

4) « *** ft) 

«a m -t q 

fill 

:goc 

•5 c w 

*o ,_ — t2i 

C w T3 

7* « C S 
. C <u ,* 

8g e S 

■a IS « 

Ef .£ a 
o .H <* a> 



ho 



O Q 

.. —I 
c O 


a. 






O -g 



5-^"5 



•S^"5 2 

£.22 e 

pS 'S. w j, 

•« C o n 
2 o * 5 

°1^ 



Si 3 S 






S .2 



M= 



c - 2 - 

"1 |t = 

ft, ,g (0 « 

* § 3.3 

« C "" 



w 



"c — J= 



<^<' 



UNDERSTANDING THE CONSUMER 57 

With unerring wisdom, they squashed it almost flat, making it a perfect frame 
for Henry VIII (see Figure 2-B). The royal tailors had, meanwhile, taken care 
to pad his already thick physique to something closely approaching a true square. 7 

It follows, that fashion trends often can be charted as they are developing, 
but sometimes the difficulty is to know exactly when there is going to be a change. 
Here, intuition must be used to supplement research, since buyers may not know 
what they are going to want next year until next year comes, and until they and 
their fellow human beings begin to react to the new products and their promotion. 
Since it sometimes takes a long lead time to develop new styles or new models 
(for example, two or three years in the automobile industry), there is much risk 
involved. It must have cost the Ford Company millions of dollars because it mis- 
judged what style preferences would be by the time the Edsel appeared. By then, 
all the advertising in the world couldn't make people want it. 

That last point underlines a more general truth. Consumers' behavior is not 
just a matter of existing wants or motives. They usually can be "sold"— that is, 
influenced to want something enough to go out and buy it or to reach for it when 
they see it on the shelf in the self-service supermarket. This is what gives mar- 
keters their dynamic opportunities. But there has to be some basis already existing 
in people's way of life or thinking, whether operative or still latent, conscious or 
just on the border of consciousness, before selling can be more than momentarily 
successful. This is what makes it necessary for marketers to be analytical, scientific, 
and in particular to take a rigorous approach to exploring and describing the com- 
plexity of people's motivations. Hence the particular need to use the scientific 
techniques of marketing research. 



MARKETING RESEARCH 

Some psychologists and marketing researchers trace buying motives back to 
sex factors, drives, guilt feelings, and so forth. This impresses many businessmen, 
and so it helps such researchers sell their services. The fact is, however, that we 
do not need to delve quite so deeply into Freud to see that people don't always 
know why they buy something or don't buy it, or are unable to explain their 
attitudes to questioners. 

For example, in selling subscriptions to a business magazine published by the 
graduate school of business of a large university, there is reason to believe that 
many businessmen subscribe to it because it makes them feel professional and 
important— they unconsciously see themselves as associated with university-trained 
executives, with presidents of big companies, and the like. It is particularly appeal- 
ing to men who are in business, but who originally wanted to be doctors or lawyers 



7 For the full study from which this material is drawn, see Dwight E. Robinson, "Fashion 
Theory and Product Design," Harvard Business Review, November-December 1958, pp. 
127-29. 



58 UNDERSTANDING THE CONSUMER 

or architects; it gives these men a feeling of dignity that they miss because, for 
one reason or another, they were not able to go on with their originally intended 
professions. 

But if businessmen are asked why they need the magazine, they say, "Because 
it helps me solve my business problems." They are not trying to invent the an- 
swer; they believe it— and presumably the magazine does serve some practical need 
on their part (beyond purely psychological motives), or they wouldn't keep on 
buying it year after year when subjected to the selling appeal of competitive maga- 
zines or services. The point is, however, that the same businessmen just do not 
realize that a strong psychological factor helps to make them prefer this particular 
kind of reading. 

Another difficulty in finding out what people feel or want in relation to prod- 
ucts is that, when surveyed or questioned, they immediately become self-conscious 
or try to give the answer that they think is expected of them. Take the problem 
of trying to predict what styles and colors of wallpaper housewives will prefer 
next year. This is not only a tough, but a very serious, problem for wallpaper 
manufacturers— they have to design patterns, prepare printing plates and presses, 
make up books of samples, send them out to dealers in advance of the buying 
season, and then sit back and hope that people will buy from them rather than 
from their competitors. If they have guessed wrong, it is too late to change. Yet 
people's ideas about color and style change from season to season— so how can 
one predict what patterns will sell? 

One manufacturer simply went out and asked people. His researchers or- 
ganized panels from members of women's clubs in churches across the country, 
offering to pay twenty-five cents per participant to their treasuries. They took great 
care to select groups so that they had a representative cross section of American 
housewives— so far, so good. Then they showed them different patterns they were 
thinking about making for the next season and asked them to vote which was 
the best. Fortunately, they didn't go by the results, for the votes were completely 
at odds with what sold well the next season. 

So they tried again, only this time they added one new wrinkle: they an- 
nounced they were going to draw lots, and the lucky winners would get a room 
papered free. Down at the bottom of the ballot was a blank, where each woman 
wrote in the pattern she wanted for her room if she won. That did reveal some 
of the trends for the next season, at least to the extent that the manufacturer 
found that the patterns that were not chosen at all were likely to be complete 
duds; that is, he could screen out the bad ones, even though he couldn't always 
catch the top successes, for the very simple reason that new colors and styles 
gather force rapidly as housewives decide they like what they see in magazines 
and on television and in other homes around them. 

Note, again, that when the housewives voted for the various patterns, they were 
trying to be honest and helpful— perhaps too helpful— for they were trying to be 
experts, trying to pick the style and color they thought others would think right. 



UNDERSTANDING THE CONSUMER 59 

When they had to choose for their own room, though, they thought of how it 
would look with their rugs and drapes and furniture, what the "man of the house" 
would say, what the neighbors would think, and so on. This is the kind of strata- 
gem that sometimes has to be used to get people to answer correctly, even when 
they do not know themselves why they think as they do. 

One of the neatest of all such stratagems was the way a researcher got at the 
reason why instant coffee wasn't selling, when it was introduced. To direct ques- 
tions, housewives answered vaguely about quality, taste, and other factors. Could 
there be a deeper reason that housewives themselves were not aware of? He showed 
two groups of housewives two different shopping lists — one with regular coffee, 
one with instant coffee, and asked which kind of housewives the list belonged to. 
Of those who saw the list with instant coffee on it, 98 per cent said: "Lazy, poor 
housekeeper," as against eight per cent of those who saw the other list. 8 They had 
revealed what unconsciously they felt. (Advertising was redirected to build up the 
good housewife connotations of better coffee, and instant coffee has become a 
great success.) 

By now, there are a host of similar devices, so-called depth interviews and 
indirect questioning, designed to get down below the surface; but they are far 
from accurate. So much depends on the marketer's having a clear idea of the in- 
formation he is seeking and on communicating his need to the researchers (which 
again means that he has analyzed or mapped out his problem). So much depends 
on the researchers' interpretations of what they think they have found. Also, by 
and large, they call for so much skill and are so time-consuming, hence so expen- 
sive, that it is difficult to get enough data to predict quantitatively— to tell how 
many people will feel this way or that way, or have this or that preference, when 
it comes time to buy. 



Quantitative vs. Qualitative 

On the other hand, there is the fact that, in many situations, simply "count- 
ing noses" is not enough. This is tied in with a distinction between quantitative 
and qualitative research— between finding answers to questions of who, what, and 
how much about consumers' behavior, and answers to the particular question of 
why— though, in actual practice, the two kinds of investigation are likely to be 
interdependent and overlapping. 

A. Edward Miller, Publisher of McCalVs magazine, has emphasized the fact 
that both kinds of research are scientific, by analogy to the work of Dr. Jonas Salk 
in developing the polio vaccine. Dr. Salk was engaging in qualitative research 
when he was injecting certain viruses into certain animals to grow the cultures 
in certain organs, which were then subject to a certain process to produce a cer- 



8 Mason Haire, "Projective Techniques in Marketing Research," Journal of Marketing, 
April 1950, p. 65. 



60 UNDERSTANDING THE CONSUMER 

tain desired vaccine. He was engaging in quantitative research when he put his 
vaccine to a nationwide field test to determine the degree of its effectiveness. With- 
out the qualitative research, there would have been no vaccine; without the quan- 
titative research, there would have been no proof. 9 

Perhaps it will be helpful in understanding qualitative research— the need 
which may arise for more than "counting noses" and the way in which such a 
need can be met— if the pertinent considerations are put together into a com- 
posite picture. Let us look at the steps a manufacturer might go through in seek- 
ing information on which to base marketing decisions— say, a soap manufacturer, 
who would have more than an average stake in consumer motivation. This is not 
an actual case; the manufacturer is many persons rolled into one, and the details 
are oversimplified. But it is realistic in the sense that it is patterned on the kind 
of problems faced in varying degree by countless companies. 

Suppose the soap manufacturer is not satisfied with his sales. In order to get 
a measure of his strength in the market, he has his marketing research director 
or some outside agency send out questionnaires to a random sample of house- 
wives—following the regular procedures of quantitative research, such as avoiding 
mention of his company's identity, so as not to bias the results, and careful check- 
ing of the mailing sample, and then of the returns, to make sure that he has a 
representative cross section, geographically and by income class. 

And suppose he finds that 30 per cent say they prefer his Brand (A), 40 per 
cent say they prefer another brand (B), and 20 per cent still another brand (C), 
with the balance divided over a score of minor brands. Normally, this breakdown 
would coincide with records of actual store sales— in which case he still wouldn't 
have made much progress toward getting helpful information. But, just to make 
the problem more difficult, suppose he has industry data showing his share of total 
national sales to be only 15 per cent, as against the 30 per cent preference figure 
from his survey— a discrepancy he might never have discovered, if he hadn't been 
fortunate enough to have some other quantitative data to check against. 

Further statistical investigation might provide some explanation. Thus, it 
might reveal that the manufacturer's Brand A has a larger percentage of prefer- 
ence than other brands in the upper-income families, who, unfortunately, do less 
home washing and hence account for fewer purchases than indicated by their 
numerical strength. Even allowing for this, however, there might remain a big gap 
between actual usage and expressed preference. 

Up to this point, the manufacturer is still "counting noses" and, with proper 
statistical controls, gaining some important information about his existing share 
of the market in relation to competitors. The same kind of surveys, repeated from 
time to time, will indicate whether he is growing in strength or the reverse— 
whether what he is doing is roughly right or wrong over-all, but with no idea 
whether the credit or blame belongs to product, packaging, price, or promotion. 



9 A. Edward Miller, "Quantifying Qualitative Research," Proceedings of the Boston Con- 
ference on Distribution (Boston, International Marketing Institute, 1961), pp. 40-41. 



UNDERSTANDING THE CONSUMER 61 



Finding Out Why 

Obviously, he needs to know the why of some of these quantitative findings. 
But it isn't enough simply to ask consumers to explain their preferences. When 
that is added to the questionnaires, the soap manufacturer finds that those who 
prefer his brand say they like it because it washes cleaner, and those who like 
other brands also say they like them because they wash cleaner! 

So he decides to send out interviewers to probe more deeply, asking con- 
sumers pertinent questions in the hope that he will get more detailed answers 
than consumers would go to the bother to give, or would be able to give in a 
written questionnaire. 

Here he begins to depart from the quantitative approach, on the ground that a 
really searching personal-interview survey on a large scale can be prohibitively 
expensive. Moreover, he may feel that all he needs is to get a picture of the kind 
of thinking consumers actually do, so as to interpret his quantitative data better; 
and for this purpose a limited number of deeper interviews with key types of 
consumers will be sufficient. (Sometimes, and particularly in the case of industrial 
goods and new products, a selective survey of thought-leaders and of potential 
users of above-average knowledge and experience can in fact come closer to the 
market and produce more significant leads than a more expensive, broader, but 
shallower survey.) 

But, again, it still isn't enough for the interviewers to ask specific questions 
such as: " Which brand of washing powder do you usually buy?" "Which brand 
do you have on your laundry shelf now?" "Which brand do you think is purest?" 
"Which brand do you think is strongest?" "Which brand do you think is most 
economical?" It might turn out that, of many consumers who consider Brand A 
the purest washing powder, quite a few fail to purchase it— or, if they say they 
purchase it, do not have a package in the kitchen at the time of the interview. 
Perhaps, also, a significant number of the same consumers indicate they feel other 
brands cost less, although actually they are in the same price range. 

The manufacturer is digging deeper, but all he has at this point are some 
new significant questions that he never was aware of before. For example, the idea 
of "purity" (which he may have been stressing in his promotion) apparently has 
some connection with indicated preference but not with actual usage— why? Ap- 
parently, other brands that are considered less pure are also considered better 
buys— why? 



Deeper Motivations 

The kind of specific questions that lead consumers to think along set paths 
and elicit one-word answers will not lead to the discovery of deeper motivations 
that the soap manufacturer now needs to uncover, if he is not to jeopardize his 



62 UNDERSTANDING THE CONSUMER 

investment of thousands or perhaps millions of dollars in the brand. By changing 
the theme of his selling approach from 'purity/' for example, he may be surren- 
dering his present appeal without replacing it with something new and stronger. 
Yet that is just the risk he would be running if he proceeded on the assumption 
that people always mean what they say. 

So the researchers proceed in a casual, conversational fashion, using provoca- 
tive leads rather than cut-and-dried questions, all with the purpose of getting the 
housewives to talk freely and with individuality— an example of the technique 
called depth interviewing. Or perhaps they show a picture that depicts the brand 
or something related to the brand, and ask the consumer to "tell a story about it," 
so they can judge from the consumer's attitude (enthusiastic or dull), as well as 
from what she happens to say spontaneously, just how she feels about the brand— 
which is the projective technique. Or, again, they may ask the consumer to match 
different brands as closely as she can against different age levels, or occupations, 
or social classes. Other techniques include word association (where the consumer 
is asked to say the first word that comes into her mind when she hears a series of 
stimulus words) and incomplete sentences (which the consumer is asked to finish). 
These techniques, and others like them, constitute the tools of what is called "mo- 
tivation research." 

On this basis, consumers may indicate that the advertising of Brand A, which 
features pictures of society matrons using it, has created the impression of over- 
elegance and association with those who can afford to be careless with money. 
And since this does not jibe with their idea of washing clothes— a tough, un- 
pleasant chore which affords them some satisfaction in being martyrs about it— 
they unconsciously react negatively and tend to reach for another brand when 
they see them all together on the supermarket shelf. 

These findings assume additional significance in the light of the fact that the 
housewives naturally feel they should have a more rational explanation of their 
failure to buy Brand A. They have seen and accepted the concept of purity which 
has been its promotional theme, for that is in line with traditional middle-class 
ideals. Pushed by their own logic, in fact, they think of it as too pure, too delicate 
for the job— the hard, workaday scrubbing job of getting the grime off their chil- 
dren's playclothes and their husbands' overalls. And so, to them, it is not as good 
a buy as other brands. 

Now, the manufacturer can better understand his quantitative data— why he 
has a relatively stronger hold on upper-income housewives, why he has more 
"preferers" than "users," and so on. And he can go back to measuring these 
qualitative factors more accurately by questionnaires or mass interviews where the 
focus is specifically pointed to bring out the differences between consumers' con- 
cepts of his brand and his competitors' brands. 

Of course, rather than going through all the steps of working up to qualita- 
tive research, as we have had the soap manufacturer do, just to illustrate why it is 
necessary to dig deeper into consumer motivations, he could have made the 



UNDERSTANDING THE CONSUMER 63 

qualitative survey first and only then have gone on to quantify the factors thus 
uncovered. And he probably would have done it this way if he were relying on 
experienced research technicians with an understanding of his problem. 10 

Deciding on Research 

Certainly, not every manufacturer or seller must always endeavor to dig so 
deeply into motivations. Far from it. Research making use of depth techniques is 
time-consuming and requires trained interviewers; at from twenty-five dollars up 
per interview, plus analysis and interpretation, a survey of this kind usually runs 
into four or five figures. 

Moreover, the problems at stake often call for information on measurable 
factors such as : number of potential users classified by geographical location; selec- 
tion, supervision, and compensation of salesmen; size of retailers' margins on this 
product compared with competing products; size of wholesalers' inventories; num- 
ber of salesmen's calls per day; size and frequency of repeat orders; circulation of 
advertising media by number and by kind of readers; and so on. 

The success or failure of a product depends on a host of such factors, all inter- 
related and all affecting the movement of the product to the store shelves, the 
showrooms, the catalogs— that is, all preliminary to putting the product where 
consumer motivations can come into play. And for many companies these more 
tangible factors constitute the area where analysis and corrective action are most 
needed, or at least will produce the most improvement for the time and money 
spent. 

Certainly, also, not every consumer motivation is always different from what 
it appears to be. In the hypothetical case about the soap manufacturer, for in- 
stance, we must not forget that consumers do want to buy washing powder in 
order to do washing. No brand can hold consumers' allegiance unless it does wash 
clean; and the surest way for any soap manufacturer to increase his share of the 
market is to offer a product that is superior in this very respect. In other words, 
not every consumer motivation is tied to Freudian subtleties and social nuances 
or— even if it is— produces effects on sales significantly different from what would 
be expected if some more ready rationalization were simply taken at face value and 
used with common sense. 

But as business becomes more complex, as operating decisions involve greater 
commitments of money, and as competition narrows the gap between products 
and thus puts a growing premium on the leverage to be secured by even small 
margins of superiority — in short, as management faces an increasingly difficult 
task, there is more and more need to be sure about initiating action and not to 
be content with policies which, however sound, could be made still more effective. 



10 See "Motivation Research — the Significance of Consumer Behavior," Acme Reporter, 
1955, No. 2. 



64 UNDERSTANDING THE CONSUMER 

Motivation research does provide more significant information. Maybe the 
soap manufacturer should change his package from a simple design in pastel colors 
(representing purity) to a bold design in darker colors (suggesting rugged 
strength ) . Also, he now would certainly know that he had better stop using testi- 
monials from society matrons; that luscious movie stars would not be much better, 
or even might be worse if they denote frivolity and high living. But he should 
gain if he builds his ads around realistic housewives— only a little neater and 
prettier than life; shown in believable suburban surroundings— just a little more 
spacious and gracious than actuality. 



Difficulty of Implementation 

The greatest difficulty with motivation research is the gap between findings 
and implementation. Consumers' concepts of products and brands are usually 
made up of a bundle of different and sometimes conflicting motivations— a verita- 
ble "bucket of worms"— much more complex than suggested in the oversimpli- 
fied case of our soap manufacturer. Even when the marketer knows why people 
buy or do not buy, he doesn't always know what to do about it. Does he try to 
remove resistances or ignore them, for example? (See Section Four.) Above all, 
he must keep in mind: 

( 1 ) That some goods are bought primarily on the rational basis of price, 
quality, and service, rather than in terms of individual feelings, aspirations, 
prejudices, and fears— though there is always some of the personal element 
present, even in the case of industrial buying. 

(2) That moving a product through the channels of trade is a complex 
process, and sales success depends on a lot more than what is involved just 
at the ultimate point where the purchase is made. 

(3) That the importance of what consumers think or can be led to 
think varies according to the product and the appropriate marketing strategy 
— ranging from being extremely crucial where the product competes for the 
mass market through eye-appeal and advertising, to being of only minor effect 
where there is a small group of large users with well-defined needs who can 
best be reached through direct salesmen. 

In short, whether a person buys or not depends on more than just what is in 
his head, his psyche, or the top of his spinal column. In a free economy, people 
have the right to buy or not, and psychologists assure us that their ability to dis- 
criminate sophisticatedly and to resist what Vance Packard has labeled the "hidden 
persuaders" is growing even faster than the capacity of manipulative techniques 
to get them to buy something for which they do not have a real or latent want. 
(There is now some evidence that people are even subconsciously able to resist 
the subconscious persuasion of "subliminal" advertising, whereby advertising mes- 
sages are delivered in the unseen split seconds between the frames of a moving 



UNDERSTANDING THE CONSUMER 65 

picture. At best, you can make observers thirsty, but not for a given brand of 
soft drink; and maybe, once people get the "hint," you cannot even do that.) 

The converse of such freedom is the right of businessmen to try to sell— so 
long as they do not engage in willful misrepresentation (as distinguished from 
enthusiastic or dramatic representation of their products, such as consumers ex- 
pect and usually know how to interpret). Cyril C. Herrmann, Vice President of 
Arthur D. Little, Inc., and Professor John B. Stewart of the Harvard Business 
School have studied consumers' propensity to take risks— whether they are cross- 
ing the street, driving a car, playing a game, or buying a new product— and, by 
experimental games designed to simulate buying conditions, they have found that 
people prefer gambling choices over nongambling choices (in their test, 93 per 
cent out of 3,600 times). Also, the further people feel they are behind in a game 
(the equivalent of feeling less than satisfied with an existing product), the bigger 
the gamble they are willing to take. 11 In other words, people are willing to inno- 
vate—to try something new even if they are not sure of it— and to this extent, 
whether they themselves realize it or not, they welcome the efforts of marketers to 
persuade them to buy new products (or products and brands which are new 
to them). 

In any event, when we try to tie people's reactions to marketers' actions, we 
need a framework for analysis (like the marketing mix, which follows) that will 
put the factors together in a pattern that will make them understandable and 
amenable to decision making. 

Marketing Mix 

Demand is a matter of interaction between sellers and buyers; or, even for- 
getting the purposeful action of sellers (which we will discuss in subsequent 
sections), between products offered for sale and buyers. At the risk of over- 
simplifying and making arbitrary distinctions between merging, overlapping fac- 
tors, we can distinguish four kinds of interaction. As a matter of fact, if one re- 
members to flesh out the skeleton of this analytical structure as he copes with 
particular problems, it will have the positive advantage of making sure that he does 
not overlook some major piece of the total picture. 

Here, then, are four categories of demand factors which, in varying com- 
binations of different degrees of each, determine whether particular people will 
buy or can be persuaded to buy particular products. (This can be conveniently 
called the "marketing mix," a term originally conceived by Professor Neil H. 
Borden of the Harvard Business School) : 

(1) The characteristics of the product itself versus what people want 

the product to be. This was the crux of the soap manufacturer's problem. 



11 Cyril S. Herrmann and John B. Stewart, "The Experimental Game," Journal of Market- 
ing, July 1957, p. 15. 



66 UNDERSTANDING THE CONSUMER 

(2) The price of the product versus the ability of people to pay it, or 
their disposition to pay it relative to the attractiveness of other uses of the 
money for other products or services. And note that, whatever their resources, 
they can stretch them by credit, by borrowing, or by installment purchases. 
This is particularly relevant in these days when most people live so far above 
the minimum level of decent living that money is not so much a limiting 
factor as it is a way of expressing priorities; and when most people are so 
sure of the stability of their future earnings that, when they anticipate them 
through credit, they really are doing no more than budgeting their income 
over time so as to enjoy next year's standard of living today. 

(3) The physical location or convenience of purchasing the product 
versus the physical accessibility of people to such "contacts," in terms of 
actual proximity and ease of transportation, and in terms of shopping habits 
(plus what they feel is signified by the kinds of outlets or sales locations 
which handle the product). Thus, fine linens will sell better in a department 
store than in a supermarket; and complex machine tools will be bought more 
readily from a technical manufacturer's representative than from an indus- 
trial or mill supply firm. On the other hand, cigarettes are carried in as many 
places as people shop for their daily wants; and standard screws and bolts 
will only be bought if they are carried in the industrial supply firm (or, usually, 
the two or three firms) that a manufacturer patronizes. 

Here also belong questions not just of the kind of wholesaler or retailer 
but of how many (whether extensive, to get the best coverage, or selective, 
to get the extra push on the part of the wholesaler or retailer who prizes his 
"franchise"); and even such further details as location in a store— upstairs or 
in the basement, in a department store; by the cash register or in a gondola 
display at the end of an aisle, or on some less prominent shelf (even what 
shelf— a high one or a low one or at eye-level), in a supermarket. 

(4) The communicability of the attractiveness of the product versus the 
receptiveness of people to such communications, depending on (a) the avail- 
ability of media to transmit the message economically, (b) the auditory or 
visual distinguishability of the product in the market (its packaging, for in- 
stance), and (c) the way people use or regard the product, being more subject 
to communication if the product means more to them either by frequency of 
use (like bread in the home or grinding wheels in the factory), by dollar 
value (like a house or a big computer), by importance of use in terms of the 
cruciality of the product in their lives or businesses (like an automobile or a 
machine tool), or in terms of its contribution to other things that are crucial 
in their lives or businesses (like baking powder in a cake that will win the 
housewife her family's approval, or the grease that will keep an important 
machine from "freezing"). 

It is the essence of marketing to reinforce or change these factors, as should 
become increasingly apparent as we go ahead. But in this section, and in the 



UNDERSTANDING THE CONSUMER 67 

cases appended to it, we are simply looking at what the elements are, trying to 
understand them, so we know what we have to work with, or build on, when we 
do consider how to take purposeful marketing action. 

Derived Demand 

For some products and sources there is also another factor, sometimes limit- 
ing, sometimes facilitating: the fact that demand is tied to the demand for other 
products and does not exist independently. Thus, photographic film bears a direct 
relationship to cameras, records to record players, and antifreeze to automobiles. 

Even in such a relationship there often is room for marketing action. Non- 
camera owners can't use film, but camera owners can be persuaded— by better 
quality, by lower price, by greater ease of purchase, by stronger promotion, or by 
some combination of these factors— to want to use more film than they otherwise 
would. 

In the case of industrial products, the scope for action may be narrower. 
Thus, it is virtually impossible to sell factories machinery they cannot use beyond 
what their volume of business requires. One of the interesting phenomena of 
marketing is the way capital equipment sales can stay steady only when the in- 
dustries they serve are expanding. If the demand for the product of an industry 
is level, the demand for production equipment will be limited to replacement 
needs and will show a declining curve— that is, unless, once again, marketing can 
convince prospective buyers that new equipment will produce more than enough 
of a saving in labor costs or quality rejections to cover its own depreciation and 
interest costs, or to step up the product's competitiveness in the market. 

Or— and for many products this is the way to expanded sales— marketing can 
persuade users to buy this product as a substitute for other products: long-distance 
telephone calls in preference to telegraph, air travel in preference to rail travel, 
a stamping machine in preference to a milling machine, refined tall oil fatty acids 
for making alkyd resins in preference to soya acids (see Case 2-9), and so on. 

In sum, each situation has to be mapped out and analyzed on its own merits, 
before there is any possibility of knowing what is the right marketing mix for it. 



LOOKING AHEAD 

The stage is now set for the sections that follow. 

In Section Three we shall round off this first stage of exploration by 
looking at problems involving the size of the potential market for a product, and 
in doing so, we will be mainly concerned with the particular characteristics of the 
product. Yet even here we shall consider a few cases where price and distribution 
outlets also have a bearing on how big the demand will be. 

Then, as we go on in Sections Four and Five to consider (a) ways of expand- 



68 UNDERSTANDING THE CONSUMER 

ing the market and (b) ways of strengthening one's position in the market, we 
shall increasingly tackle the need for making decisions where two, three, or four 
different demand factors are at stake simultaneously. Finally, in Section Six, we 
shall get into strategy and planning, and here the essence is to combine the four 
factors— product characteristics, price, physical location, and communication— in 
a purposeful pattern that will bring the company long-run success. 

Thus, we might have some farmers paying a high price for a small quantity 
of a particular pharmaceutical to treat a serious disease in their chickens because 
their feed dealers tell them the manufacturer has a reputation for leading the 
way in the scientific characteristics of his preparations. At the same time, other 
farmers (or the same farmers under different circumstances) are buying larger 
quantities of a competing brand because it sells for less and is strongly advertised. 
In the first case, the company is following a policy of basing its marketing strength 
on a combination of scientific research and high margins for dealers, while the 
second company is basing its marketing strength on large sales volume at a low 
per-unit profit. 

Note that both of the companies just described use a combination of two of 
the four major kinds of demand factors as their principal bid for the market— 
and it is a different combination in each case. It should be observed, also, that 
the same two companies must take the remaining factors into account, too; like 
every marketer, they must have a full mix of all the factors— but, again, like every 
marketer, they must emphasize one or two as their main approach. 

Sometimes this will happen by accident— by circumstance and the unplanned 
interaction of sellers and buyers. Or, as will be demonstrated later, the oppor- 
tunities can be analyzed and a logical plan of action developed. 



In the cases that follow we shall have an opportunity to try to identify the 
different demand factors that apply in different product-market situations. 



UNDERSTANDING THE CONSUMER 



69 



CASE 2-1 

Ball Brothers 



BACKGROUND 

In 1954, Ball Brothers acquired the Hy- 
geia Nursing Bottle Company as a part of 
a diversification. 12 Although Hygeia had 
produced nursing bottles for approximately 
sixty years, it sold the bulk of its produc- 
tion to hospitals, leaving the consumer 
market virtually untapped. Since Ball 
Brothers had a retail store sales organiza- 
tion of sufficient size and experience to 
handle a new product, there was no reason 
to suspect that Ball Brothers couldn't ob- 
tain adequate distribution of the Hygeia 
nurser at the retail level. The real prob- 
lem was to develop a profitable promo- 
tional program that would insure move- 
ment out of the retail stores. 

Several factors complicated the develop- 
ment of a successful marketing program: 

1. A single brand (hereafter called Brand 
"A") had 80 per cent of the consumer 
segment of the nursing bottle market. 
Consumer surveys showed that users of 
this brand were reasonably satisfied with 
the product; indeed, Brand "A" had suc- 
cessfully resisted all attempts of other 
competitors to erode its position during 
the postwar period. 

2. Hygeia's nipple and bottle were not in- 
terchangeable with the nipples and bot- 
tles of other brands. Since two out of 
three expectant mothers have had other 
children, they tend to have at least a 
small inventory of nursing bottles left 
over from these children. 



12 This case was adapted from an account by 
Charles G. Collard, Director of Market Re- 
search, Ball Brothers, in Henry Brenner, Edi- 
tor, Marketing Research Pays Oil (Pleasant- 
ville, N.Y., Printers' Ink Books, 1955), pp. 
11-15. 



3. Repeat purchases by expectant or new 
mothers could not be expected until 
they had another child. 

4. Very few consumers were aware of the 
Hygeia name. 

On the other hand, the Hygeia product 
was improved and streamlined in prepara- 
tion for invading the consumer market, and 
this was expected to aid the development 
of a successful marketing program. The 
general nature of the distinguishing char- 
acteristics that should help gain consumer 
acceptance of the Hygeia brand were as 
follows : 

1. The nipple had a wider base and a 
shorter tip which provided a rubber 
cushion for the child to nurse against. 
Hence, the structural shape of the nip- 
ple tended to be more breast-shaped 
than any other nipple on the market. 

2. The bottle had a wider mouth and a 
tapered design that eliminated shoulders 
at the top of the bottle. Hence, the bot- 
tle was easier to fill and to clean than 
any other bottle on the market. 



RESEARCH METHOD 

Consumer research was conducted along 
the following lines: 

1. The basic plan involved a direct-mail 
campaign to a list of expectant mothers, 
the latter being obtained from cooperat- 
ing diaper services. 

2. Advertising copy describing both Hygeia 
and Brand "A" was mailed to these ex- 
pectant mothers. After the mailings, the 



70 



UNDERSTANDING THE CONSUMER 



expectant mothers were asked to fill out 
a questionnaire pertaining to their ex- 
perience with and attitudes toward 
nursing bottles. In return for their co- 
operation, respondents were offered a 
free set of either Hygeia or Brand "A" 
nursers, whichever brand they preferred. 
Hence, some measure of the relative 
effect of alternative advertising copy ap- 
proaches was provided, as well as some 
indication of Hygeia's relative position 
vis-a-vis Brand "A." 
3. Finally, a sample Hygeia nurser and a 
sample Brand "A" nurser were mailed 
to the expectant mothers to provide 
them with an opportunity of inspecting 
each brand. The purpose of this step 
was to determine whether the differ- 
ences in physical characteristics of the 
two products were discernible and of 
interest to the potential nursing bottle 
purchasers, and whether this sampling 
operation did cause any switching of 
choices from Brand "A" to Hygeia, and 
vice versa. 



RESEARCH RESULTS 

A total of 1 59 expectant mothers in one 
area were used for the experiment. The 
researchers felt that the results were of 



such a dramatic nature that a larger sam- 
ple was not needed. 

After the first step, in which expectant 
mothers were mailed advertising copy for 
both brands, 29 per cent of the respond- 
ents selected Hygeia, 66 indicated prefer- 
ence for the established Brand "A" nurs- 
ers, and 5 per cent indicated no particular 
preference for either brand. However, after 
the second step, in which respondents 
were mailed a sample of each product, the 
results were sharply reversed: 65 per cent 
now chose Hygeia, 33 per cent continued 
to choose the entrenched Brand "A," and 
2 per cent didn't care what brand they 
received. None of the original Hygeia re- 
sponses switched to Brand "A." Further- 
more, while only 20 per cent of those who 
had had previous firsthand experience 
with Brand "A" chose Hygeia on the first 
wave, 63 per cent chose Hygeia after hav- 
ing the opportunity to inspect the product. 



Can Ball Brothers count on a strong 
demand for the Hygeia bottle? What de- 
mand factors are involved here, and which 
one seems most crucial — product charac- 
teristics, location of sale, price, commu- 
nicability of appeal? What problems, if 
any r does management have to solve if the 
product is to be successful? 



UNDERSTANDING THE CONSUMER 



71 



CASE 2-2 

Farmcrafter, Inc. 



In 1961, Farmcrafter, Inc., was consider- 
ing the manufacture and sale of a "do-it- 
yourself" repair kit for farmers, whereby 
they could make many kinds of repairs 
to their farm machinery without having to 
interrupt their chores to take the machin- 
ery to the city. 

To test the reaction of the market to 
this new product, the company mailed five 
different offers, at five different prices, to 
10,000 farmers. The design was so drawn 
as to produce five mailings of 2,000 each 
specifying one of the five prices, and di- 
vided into five lots of 400 each making 
one of the five offers. Thus there were 
2,000 of each offer as well as each price. 

The offers were: (a) to send further in- 
formation to those returning the coupon, 
(b) same as a, except the coupon was to 
be accompanied by a dime to cover mail- 
ing costs, (c) to send the name of the near- 
est hardware store carrying the repair kit 
to those returning the coupon, (d) the 
same as c, except the name to be sent was 
that of the nearest farm equipment dealer, 
and (e) to send a salesman to call. 13 

The results of the mailing, which de- 
scribed the repair kit in glowing terms and 

13 The company of course planned to send 
an appropriate notice to everyone who sent 
in a coupon, including refund of the dime, in 
case it decided not to go ahead with the 
product. 



illustrated it in color, are shown at the bot- 
tom of the page (percentage of coupons 
returned). 

The reactions to the five prices were 
also analyzed by size of farms (percentage 
of coupons returned) : 

DIFFERENT SIZE FARMS 



Price 


Large 


Medium 


Small 


Average 


$75.00 


25% 


36% 


8% 


23% 


$65.00 


27 


38 


7 


24 


$55.00 


26 


34 


12 


24 


$49.75 


25 


48 


20 


31 


$45.00 


27 


51 


18 


32 



Average 26 



42 



13 



27 



The company had available this infor- 
mation about the number of farms of 
different sizes: 



Under 100 acres (small) 
100 to 500 acres (medium) 
Over 500 acres (large) 

Total 



2,561,185 

1,899,053 

322,178 

4,782,416' 



What inferences should the company 
draw about the nature and strength of de- 
mand for its proposed repair kit? What do 
you think of this way of testing reaction to 
a proposed new product? 



14 Statistical Abstract of the United States: 
J959, p. 620. 



DIFFERENT OFFERS 





(a) 


(b) 


(c) 


(d) 


(e) 






Infor- 


Info. 


Hardware 


Equipment 


Salesman 




Price 


mation 


for 10? 


store 


dealer 


to call 


Avera 


$75.00 


42% 


11% 


23% 


32% 


6% 


23% 


$65.00 


39 


14 


25 


31 


9 


24 


$55.00 


43 


12 


25 


34 


7 


24 


$49.75 


55 


19 


35 


43 


12 


31 


$45.00 


52 


20 


33 


41 


13 


32 


Average 


46 


15* 


28 


36 


9 


27 



'But 2% failed to enclose a dime. 



72 



UNDERSTANDING THE CONSUMER 



CASE 2-3 

Schmidt Packing Company 



THE PROBLEM 

In January 1959, Herr Schmidt, presi- 
dent of the Schmidt Packing Company, 
was considering the information he had 
gathered on "Brown and Serve" pork sau- 
sages, preparatory to deciding whether the 
company should produce and market such 
a product. 

THE COMPANY 

The Schmidt Packing Company was a 
regional packer and marketer of pork prod- 
ucts. The packing plant was located in 
Frankfurt, Germany. The markets the com- 
pany served included all of Hesse, a small 
part of Wurttemberg, Baden, and Munich. 
The company maintained two sales offices, 
one in Mannheim and the other in Mu- 
nich. Each sales office was the headquarters 
for approximately 20 salesmen who sold 
Schmidt products to grocery outlets. In 
1959, the company had had a net after-tax 
profit of 1,260,000 DM ($300,000) on 
sales of 315,000 DM ($75,000,000). 

THE PRODUCT 

According to the salesmen of the sausage 
machinery Schmidt used, "brown and 
serve" sausage had been introduced in 
the United States with great success by 
the Fisher Company, an American meat 
packer, in 1958. Essentially, the brown 
and serve sausage was a sausage that was 
thoroughly precooked at the packing plant 
and therefore required only three minutes 
in the skillet of the housewife in order to 
become hot and brown, in contrast to the 



thirty minutes or more required to cook 
the standard sausage and make it ready 
for the table. Actually, the brown and 
serve sausage could be eaten safely directly 
from the package; the frying served only 
to heat and brown the product and thus 
make it more appetizing and eye-appeal- 
ing. 

During the cooking process, whether 
done at the packing plant or in the home, 
approximately 50 per cent of a raw sausage 
was cooked away. Therefore, in order to 
present the housewife with precooked sau- 
sages in a familiar size, the Fisher Com- 
pany management had considered it desir- 
able to make them twice as large as normal 
in the uncooked state and then cook them 
down to the standard size. Fisher also 
priced the product at twice the cost of the 
same weight of standard sausage. 



REQUIRED INVESTMENT 

After some investigation Herr Schmidt 
had discovered that the process of making 
brown and serve sausages involved, in addi- 
tion to new machinery for making the 
sausages twice as large, a surprising num- 
ber of cooking, transfer, and temperature 
change operations, and that components 
of equipment to make the process efficient 
on a production-line basis would cost in the 
neighborhood of 840,000 DM ($200,000). 

Herr Schmidt believed that a six-month 
introductory program for the areas he 
wished to cover — involving the hiring of 
demonstrators along with newspaper and 
radio advertising — would cost approxi- 
mately 168,000 DM ($40,000). 



UNDERSTANDING THE CONSUMER 



73 



FINANCE 

The company's financial position was 
such that the company could not go into 
the production and distribution of pre- 
cooked sausages without resorting to a 
bank loan. On the other hand, its credit 
was good, and Herr Schmidt did not be- 
lieve that there would be any difficulty in 
obtaining a loan on a long-term basis. 

PROSPECTS FOR THE PRODUCT 

As far as Herr Schmidt could ascertain, 
the only experience with brown and serve 
sausages had been that of the Fisher Com- 
pany. Herr Schmidt asked his advertising 
agency people if they knew of any surveys 
on consumer response to a brown and serve 
sausage, or if they had any other material on 
consumer food preferences or food buying 
habits that might be helpful. All they could 
supply, however, was the information in 
Table 2-C. 

At the suggestion of the U.S. Agricul- 
tural Attache, Herr Schmidt also contacted 
the U.S. Department of Agriculture, which 
often contracted with independent agen- 
cies to run consumer preference tests on 
new products in which the department was 
interested. These tests usually involved 
getting the product on the shelves of a 
supermarket. Then an interviewer at the 
store would ask purchasers to fill out an 
address card to facilitate later follow-up 
with a questionnaire. However, the Agri- 
culture Department had no data available. 

Herr Schmidt's next move was to write 
to an acquaintance in the United States 
who was a member of the American Mar- 
keting Association and knew many mar- 
keting research people in American com- 
panies, to see whether he could glean any 
crumbs of Fisher's experience with the 
brown and serve sausage. He was very sur- 
prised to receive a considerable amount of 
information in a letter from the Associate 
Director of Marketing Research for the 



Fisher Company, in response to a request 
from his friend. Pertinent parts of the let- 
ter are reproduced below: 

Once the idea of precooked sausage was 
created, our lab began the long process of 
testing, tasting, and improving the prod- 
uct. The appearance of the sausage, the 
seasoning, the length of precooked time, 
all had to be considered and perfected by 
the lab's specialists until the product ap- 
peared to meet consumers' specifications. 
When the experimental stage was com- 
pleted, we then contacted homemakers for 
their reactions. 

The first step of the brown and serve 
consumer testing program began with a 
home preparation test. The purpose was 
principally exploratory. Information was re- 
corded from the interviewers' observations 
on problems encountered by consumers on 
such points as package opening, instruc- 
tions for preparation, degree of brownness, 
and taste comments. The initial reaction 
to the proposed product was extremely 
favorable. 

The second phase of the consumer test- 
ing was begun a week or so later. The pur- 
pose was to measure consumers' degree of 
acceptance for the fully cooked pork sau- 
sages. Two approaches were used in deter- 
mining consumer acceptance of the brown 
and serve. About half the consumers rated 
the product on how well they liked it. The 
balance of those tested appraised it from 
the standpoint of seasoning. Consumers 
were not informed that the sausages were 
the fully cooked type. From the question- 
naire, which was designed to obtain the 
degree of general acceptance, about two- 
thirds of the participants liked the product 
very much. Of those asked to appraise the 
seasoning, almost 30 per cent felt that 
the product was "too highly seasoned" and 
64 per cent thought it "just right." 

Several consumer home tests followed 
comparing brown and serve with the stand- 
ard fresh product. The variable was the de- 
gree of freshness in both products. Both of 
these tests again indicated the favorable 
attitude toward the precooked variety. 

A sales test was set up on a pilot scale, 
using a small midwest town to obtain some 



74 



UNDERSTANDING THE CONSUMER 



preliminary indications of consumer ac- 
ceptance of this new product. Interviews 
were made within the store following three 
weeks of demonstration and promotion of 
brown and serve sausages. Name recogni- 
tion and repeat buying were found to be 
very favorable. The penetration of the mar- 
ket and retention of buyers of this brand 
were checked again six months following 
the store test by making a house-to-house 
survey. An expansion to the large area 
around the original city was made. Sales 
performance was measured on a store-to- 
store basis. 

With the results favorable at every point, 
Fisher followed an orderly procedure in 



advancing the promotion of brown and 
serve from market to market to the extent 
that our facilities would permit. Spot 
checks on consumer reaction followed in 
several markets as we progressed. 



What can Herr Schmidt glean from 
Fisher's experience and the data in Table 
2-C? Why should (or should not) demand 
for "brown and serve" sausage be different 
in West Germany from what it is in the 
United States? 



PURCHASES OF PRE- PACKED, CANNED AND FROZEN FOODS, 1956 
(Percent of Purchases of Each Type of Food) 





"o 
O 


a) 
H 


u 
a 


& 


u 
o 


i 

u . 

3* 


u 
1 


01 

■a 

ei 

H 

s 

u 
a 


a 
o 
o 
o 
O 


8 
0< 


[W. GERMANY | 










Never Buy- 


3 


10 


1 


4 


8 


3 


11 


31 


6 


19 


Buy Pre-Packaged 


80 


81 


43 


45 


67 


69 


84 


53 


85 


41 


Packaged by Retailer 
after sale 


16 


8 


58 


51 


28 


30 


5 


16 


8 


39 
























| AUSTRIA | 








Never Buy 


29 


8 




3 


8 


14 


10 


48 


11 


40 


Buy Pre-Packaged 


26 


76 




21 


48 


74 


71 


35 


77 


8 


Packaged by Retailer 
after sale 


46 


18 




77 


45 


13 


20 


18 


8 


52 



IMPULSE FOOD BUYING, 1956 
WHY DID YOU BUY IT? 

In percentage of those who bought at impulse 





Austria 


West 
Germany 


Italy 


Nether- 
lands 


Norway 


Display reminded housewife 


48 


39 


51 


37 


25 


"Momentary appetite" 


25 


18 


19 


20 


- 


Bargain, low price, sales discount 


13 


20 


7 


19 


4 


Sales staff drew attention 


12 


8 


6 


4 


5 


Reference to sales material 


3 


16 


- 


4 


1 


New product, bought to try 


4 


3 


6 


2 


5 


Scarce product 


- 


- 


1 


- 


10 


Other reasons 


- 


- 


8 


9 


- 



Table 2-C 



UNDERSTANDING THE CONSUMER 



75 



CASE 2-4 

Bender Mattress Company 



The Bender Mattress Company of Los 
Angeles, California, manufactured mat- 
tresses, box springs, convertible sofas, and 
some furniture. The greatest portion of its 
business was in mattresses and box springs, 
under the brand name Restful. It sold to 
retail stores in Southern California and 
the territory up to but not including San 
Francisco. 

The major concern of the company was 
to increase unit sales. In this industry, ac- 
cording to Bender executives, prices are 
predetermined by competition, particu- 
larly in national companies, and average 
costs are approximately constant through- 
out the significant volume range. Thus, 
increased profits become a function of in- 
creased volume, which, in turn, is de- 
pendent largely upon the effectiveness of 
marketing efforts. 

In 1961, 60 per cent ($150,000) of the 
Restful advertising budget was being 
spent to sponsor the 11:00 P.M. newscast 
on Tuesday, Thursday, and Sunday over 
Channel 4; 24 per cent ($60,000) in co- 



operative newspaper advertising; and 16 
per cent ($30,000) on point-of-sale ma- 
terial. This media plan had been in effect 
since January, 1959. 

For Downy, a competing brand, the 
total amount and breakdown of advertis- 
ing were estimated to be approximately 
the same. The Nighttime Company, in 
contrast, was spending about as much but 
concentrating it on cooperative advertis- 
ing in newspapers. Restful and Downy 
were the only two brands using TV. 

In 1961, questionnaires were mailed to 
a sample of 6,000 family units in the Los 
Angeles metropolitan area selected at ran- 
dom; and 999 usable returns were re- 
ceived. Some of the resulting tabulations 
follow. 



What can the Bender management 
learn from the tabulations as to the effect 
of various factors on the purchase of the 
brand of mattress? Should it make any 
changes in its promotional mix? 



TABLE 2-D. BRAND DECISION- 
ALL RESPONDENTS 



TABLE 2-E. BRAND RECOGNITION - 
ALL RESPONDENTS 



Decision on brand to purchase made by: 

Husband 

Wife 

Both 

Do not remember 

No response 

Total 





Brand Name 


9% 


Night-time 


27 


Restful 


46 


Downy 


7 


Bye -Bye 


11 


Californian 


00% 


Posture-plus 


Sleep-Rest 




Angelic 




Moonlight 




Other 




No response 



90% 
88 
87 
82 
65 
34 
25 
16 
10 
12 
3 



TABLE 2-F. 



(a) PRE-PURCHASE PREFERENCES: ALL RESPONDENTS 



Restful 

Bye -Bye 

Downy 

Night-time 

Calif ornian 

All other 

No preference 



2 

6 

10 

2 

8 
65 

100% 



(b) BRAND PURCHASED AND BRAND PREFERRED: RESPONDENTS WITH 
PRE-PURCHASE PREFERENCE 



Pre -Purchase Brand Preference 

















All 


Brand Purchased 


Restful 


Bye -Bye 


Downy 


Night-time 


Calif ornian 


Others 


Restful 




77% 


- 


- 


6% 


- 


- 


Bye-Bye 




- 


82% 


- 


1 


- 


- 


Downy 




2 


6 


83% 


- 


- 


5% 


Night-time 




12 


6 


- 


82 


- 


3 


Calif ornian 




2 


- 


2 


1 


82% 


1 


All others 




2 


6 


6 


6 


12 


84 


Do not reme 


mber 


5 


- 


9 


4 


6 


7 


Total 




100 


100 


100 


100 


100 


100 



(c) RECALL OF ADVERTISING: ALL RESPONDENTS 



Recalled Brand Advertised 





Recc 


tiled 














Brand Not 




Mattress 














Recalled or 


Source of 


Advertising 




Bye- 




Night- 




All 


Not Related 


Advertising 


Yes 


No 


Restful Bye 


Dotvny 


time 


Calif ornian 


Others 


to Source Total 


Ratio 


16% 


84% 


2% 


1% 


2% 


2% 


* 


* 


9% 16% 


TV 


76 


24 


26 


8 


31 


4 


1 


* 


6 76 


Newspapers 


48 


52 


6 


2 


4 


6 


1 


3 


26 48 


Magazines 


44 


56 


5 


3 


3 


9 


2 


* 


22 44 


Billboards 


15 

i 1%. 


85 


1 


1 


1 


1 


1 


* 


10 15 


*Less thar 





(d) RECENT AND NON-RECENT PURCHASES BY PRE-PURCHASE BRAND 
PREFERENCE: TV VIEWERS ONLY 



Pre -Pur chase Brand 
Preference 

Restful 
Bye -Bye 
Downy 
Night-time 
Californian 
All others 

Total 



Recent Purchasers 


Number 


Per Cent 


23 


28 


3 


4 


18 


21 


15 


18 


3 


4 


21 


25 



83 



100 



m- Recent Purchasers 


Number 


Per Cent 


27 


19 


10 


7 


21 


15 


43 


31 


12 


8 


28 


20 


141 


100 



76 



oo 

3 p 



en n m to 
h n cm no 



b« teg 

NOlO) <=>\CD 
i-H .-H CM TO 



a* 








co 


eo 


CM 


c- 


cm 


*"• 


CM 


eo 



eR 



we 


e* lb 4 * 


co r- co O 


CD — « CO O O 


t* cm TO 


HHC<5 TO 



8 0) 

■*- . s 

o r o 

2 ^ A 

^ "2 «? 

O g 03 

q £ ft: 



b* 

T~l ,-1 T O 

cm cm to 



b* tefc 

CO CO i-l co o 
r-l CO TO 



teft 

OS C~ O O 

cm m o 



b*^ eS 

CO CD O «-< O 

^-t ^h cm m o 



CD CO IT) CD O 

HHPJ TO 



b^ b* 

CM CO T iH O 
CM CM CM CO O 



CO CD CD O O 
i-H CO TO 



O CM CO LO O 
CO H CM CO O 



CO CM O O O 

H i-H CO T O 



b* teg 

r- o> cd oo o 
i-i co co o 



PQ 
O 

I 

u 

w 

i'| 

w 9 

w ^ 



s 




s 

1 


bR |e£ 

•H CO i-i O O 
CM CM CM CO O 






0) 

0) 


b^ leR 

noicohO 
cm cm h co o 



cd o as in 

CM CM T 



b* 

in CD CO CO 

CO H H CO 



b^ 








as 




in 


in 


CM 


,H 


CM 


n 



'■& 


e£ leR 


b^ 


b^ 




co as as t o 


CO H CO CO 


o 


"t5 


i-l tH CM CO O 


HNNN 


o 


s 


IrH 




1— ( 



s* lbs 

CD CO O i-l O 
H CM CO CO O 



b^ b^ 

T 00 00 O O 
CM CM i-H CO O 



bS b^ 

(OHOfflo 

CM CM CM CO O 



b^ b^ 

O T CO CO O 
CO i-l CM CO O 



b^ b* 

CO as CM CD o 
•H CO TO 



bs teg 

co in t- cm o 

rH ^H CM TO 



6 



b^ |a£ 

t cm m as o 

t-< i-l CO CM O 



tes 

os in in o 

CO TO 



b^ te^ 

co as o mo 
CM .-h co cm o 



•fe* 


T CM T O O 


oo 


CM CM CM CO O 


ft) 


1 T-l 



b^ b^ 

as c- co t-i o 

i-H CM CM CO O 





K 






ft) 


_o 






a 








•*« 


■C» 






■v* 


00 






,0 


ft) *j 






"5 ft) 






1 * 


S "S 


oo Q< 


a 


o o 


e o 


s £ 




a a-. 


oo C 


Q "* 


3 
u 


C 00 CTl 




S >, 


Q 


•- ■ j_ o 


2^ 


a 


■K « H 


Cj 


£ 


o o 
55 £ 


fe. 


Si 







2 £ g 

oo B r; s w rt 

> . o - ^ o 

§ u £ 5 5 



b§ tes 

t-om oo o 

HNN CO O 



b^ IS* 

CM O t~ I-4IO 

^-i ^ CO TO 



ft) £ 



£ | 



£J co 

h c 

coo 
ctj a a— i 
r: P ■ a 

Sh C (I, <_> 

>> O — tn o 
CD C O O 

> S 2 2 



h 

S -^ 

£ a 



ft) > 



>> o 



u c 

o o 

a a—, 

£ S^ 



Ih c 

o o 



— r: p « bb 



ft) O 

5 > 



IS 2* 



77 



78 UNDERSTANDING THE CONSUMER 



TABLE 2-H 
SHOPPING HABITS-ALL RESPONDENTS 

Item 



Number of stores shopped 




One 


33% 


More than one 


57 


Do not remember and 




No response 


10 


Total 


100% 


Furniture benight with mattress 




Bedroom set 


23% 


Springs and bed 


16 


Springs 


28 


Purchased mattress only 


28 


Other and no response 


5 


Total 


100% 


Type of store in which mattress 




was purchased 




Department 


31% 


Furniture 


36 


Discount 


4 


Sleep shop 


3 


Other and no response 


26 


Total 


100% 



TABLE 2-1. 

FACTORS RELATED TO SELECTION OF STORE WHERE 
PURCHASED- ALL RESPONDENTS 

Relative Importance 





Very 




Not 


No 




Factors 


Important 


Important 


Important 


Response 


Total 


Location of store 


14% 


16% 


33% 


37% 


100% 


Reputation of store 


54 


19 


4 


23 


100 


Generally shop there 


20 


12 


29 


39 


100 


Store advertising 


9 


15 


32 


44 


100 


Special price 


31 


17 


18 


34 


100 


Store salesman 


13 


13 


31 


43 


100 



UNDERSTANDING THE CONSUMER 



79 



CASE 2-5 

Prune Research (A) 



In June, 1952, the members of the Ad- 
vertising and Trade Promotion Subcom- 
mittee of the California Prune Advisory 
Board were considering what action they 
should take on the basis of a progress re- 
port recently received from the Institute 
for Research in Mass Motivations, Inc. 

The California Prune Advisory Board 
consisted of members representing 18 
prune processors and some 6,500 growers. 
The California Prune Marketing Program, 
which was administered by the board, was 
financed by an assessment of $1.50 per ton 
on grower members and an additional 
$1.50 per ton on processor members. 15 

While the California Prune Advisory 
Board had spent only $15,000 on advertis- 
ing during the preceding crop year, it esti- 
mated that, with its recently increased 
assessment rate, approximately $200,000 
would be available for advertising during 
the 1952-1953 marketing season. 

Export markets for U.S. prunes had 
practically disappeared during the period 
following World War II. Experts believed 
that only about 15,000 tons could be sold 
abroad annually through free trade chan- 
nels without subsidy assistance. During 
1938, over 100,000 tons of prunes had 
been sold to foreign markets. Annual do- 
mestic consumption, on the other hand, 
had, over the past 20 years, remained in 
the neighborhod of 115,000 tons despite 
the considerable population growth that 
had taken place during that period. 

The prune industry had participated in 

15 This total assessment of $3.00 per ton ul- 
timately resulted in revenues of $361,559 on 
120,520 assessable tons for the August 1, 
1952 to July 31, 1953 crop year. 



two major marketing research efforts. Since 
October 1, 1949, the Market Research Cor- 
poration of America had been keeping a 
running record of domestic consumption of 
prunes and prune juice by means of a na- 
tional panel of 4,300 household consumers. 
MRCA provided monthly, quarterly, and 
annual reports on volume of prunes pur- 
chased by regions and type of outlet, per 
buying family, prices paid, and other fac- 
tors. This project was sponsored jointly by 
the citrus, prune, and date industries, and 
the U.S. Department of Agriculture's Re- 
search Marketing Administration. Industry 
and government each paid 50 per cent of 
the amount charged for the service. The 
annual cost to the prune industry for its 
share was $12,200. 

The second marketing research project 
was to hire the Institute for Research in 
Mass Motivations to determine what atti- 
tudes toward prunes then existed in the 
minds of consumers and nonusers. The 
board and its advertising agency wanted 
this information as an aid in formulating 
an advertising campaign. Two hundred 
depth interviews would be conducted, and 
the project would cost $7,500. 

The Institute for Research in Mass Mo- 
tivation, at the end of a month's work, 
made a progress report, which follows in 
abridged form: 



PROGRESS REPORT 

This preliminary report outline is based 
upon the analysis of about eighty depth in- 
terviews comprising a random sample of 
respondents. 



80 



UNDERSTANDING THE CONSUMER 



In addition, we have availed ourselves of 
the information obtained through observa- 
tions we have conducted in a number of 
relevant localities. These comprise homes, 
restaurants, cafeterias, chain stores, inde- 
pendent grocery stores, and fruit stores. 

Studies conducted by our organization 
dealing with related fields have given us 
additional insight into the problems faced 
by the prune industry. Particularly, the 
knowledge we have derived from our stud- 
ies on people's attitudes toward margarine 
and canned foods has helped us consid- 
erably in understanding the specific prob- 
lems of prunes. We have also conducted a 
brief survey of the available literature on 
fruit and food in general. 

The following pages are a brief outline 
of the present structure of this study. A 
number of conclusions have been drawn 
on the basis of our research findings to this 
date. Further research will contribute ad- 
ditional information and amplify this pic- 
ture. The findings reported here are by no 
means final. Rather, they should be con- 
sidered as an indication of the over-all 
direction that we are pursuing. 

People Like to "Like" Prunes 

We listened to American consumers of 
all ages and walks of life, and studied their 
comprehensive responses to the subject of 
prunes. What struck us first was the flood 
of complaints and accusations directed at 
a product of nature and American agricul- 
ture which, one should assume, does not 
deserve so much antagonism. . . . 

Now why should prunes of all foods be 
so heartily disliked? There are many foods 
with much more controversial qualities, 
products as sour as lemons or as bitter as 
beer, as slimy as oysters or as "strong" as 
blue cheeses, all of which are generally re- 
spected and do not draw such a torrent of 
abusive language upon themselves. 

We know from many practical experi- 
ences as well as from scores of laboratory 



experiments that "taste" and "flavor" are 
something highly subjective and change- 
able and that, within limits, foods are in- 
clined to live up to the flavor which the 
eater has been promised, if the promise 
has been made in an "appealing" and 
catching way. In other words, there is 
hardly anything that could not be made 
"tasty" and acceptable by appropriate pro- 
motion. 

On the other hand, we noticed that 
many respondents admitted eating prunes 
"when they are around," even liking their 
taste, and in their next sentence they would 
frown upon them, or disparage them; they 
gave the impression of being ashamed of 
having any relations with prunes. 

Such and similar observations convinced 
us, more and more, that people deep down 
do not dislike prunes, no matter what they 
say. Rather, they are prevented from lik- 
ing them by their own and their environ- 
ment's biased attitudes. Most eaters, we 
believe, would feel relieved if, in some 
way or other, they would get a chance to 
accept the prune and to enjoy eating it, 
just as they enjoy other fruits. We feel 
that they are somehow waiting to be 
shown ways and means to "meet" the 
prune. 

But such "negative reactions," we real- 
ized, are not just accidents. They cannot 
be caused by objective facts either. They 
must be due to deeper-lying causes, causes 
of an emotional nature. Such emotional in- 
terferences with the perception of facts are 
often unconscious mechanisms. Presenta- 
tion of facts alone usually does not correct 
such biases. 

Six Main Prune Biases 

We have found so far that there are at 
least six main biases directed against 
prunes and prune products, constantly hin- 
dering the acceptance of the prune. [These 
are illustrated with quotations from the 
interviews.] 



UNDERSTANDING THE CONSUMER 



81 



A. Prunes Are Despised as Symbols of 
Sterile Old Age 

"I hate the look of stewed prunes — lying 
sort of dead in their juice, falling apart, 
and ripped-up looking. There is something 
revolting about the look of them. I would 
never dream of eating a stewed prune now. 
The only way I like them is raw. I also 
wouldn't buy anything cooked that had 
prune in it — prune cake or anything of 
that sort. For some reason it is just un- 
pleasant to me." 

"I actually don't know what kind of 
people eat prunes — I do know that I would 
be surprised to find that any of my friends 
served them cooked. I somehow think of 
them as middle-aged and from the old 
country. I see them in ugly old cafeterias, 
the juice looking sort of wan and tired. 
And then there is that old slogan — she's a 
prune — meaning dried up, puritanical, 
dull, uninteresting." 

B. Prunes Are Suspected of Being Devital- 
ized, Denaturized 

"A plum is a fresh, juicy and meaty 
fruit, while a prune is dehydrated, shriveled 
up, leathery, unappetizing, something that 
I would not look at if it weren't for cer- 
tain digestive values they seem to have." 

C. Prunes Are Resented as a Laxative 
"Prunes are definitely a laxative, a medi- 
cine. I don't particularly care for their 
taste, but I eat them from time to time 
when my digestion needs a push. I some- 
times order them in a restaurant on trips 
since, when traveling, my digestion seems 
to be more haywire than usual." 

D. Prunes Are Disliked as a Symbol of 
Parental Authority 

"I never eat them. I never have them in 
my house. When I was a kid, my mother 
used to give them to me, but I never cared 
for them. She made me eat them because 
they were good for us. Sometimes she gave 
them to us as a physic; she'd stew them 



with lemon and we'd have them for desert." 
'The first thing I think about is a laxa- 
tive action and the way my roommate used 
to insist always that we eat our prunes and 
bran to keep regular. It didn't bother me. 
I liked them so I ate them. She had quite 
a bit of trouble with laxatives and she had 
to take bulk food all the time. That's why 
I think of laxative action with prunes all 
the time, because of her. I don't think 
I've had a prune since I stopped living 
with her — yes I did once at my mother-in- 
law's. I thought Shirley would be pretty 
pleased with me. They are just a food that 
I don't think about. Like when I am shop- 
ping I just don't think about buying 
them." 

"I recall prunes as one of the unpleasant 
desserts forced on us at home during the 
time I was growing up. It was very distaste- 
ful to me. They were eaten in my home 
for health reasons. Prunes were considered 
healthy by my family, but were terribly 
prepared — a little sugar, a little cinnamon 
and a great deal of lemon. When I got to 
the age of reason, I refused to eat them 
and didn't eat them for 10 years. I began 
to eat them again when I saw a recipe for 
prune and cottage cheese salad, and I 
started eating that once in awhile. Then, 
when I was pregnant I ate prunes a lot 
between meals, raw because I thought they 
were low calorie, but I found out later that 
they're actually high in calories. Then, 
with Laura, my attitude toward prunes 
changed because she was always consti- 
pated as a baby and prunes were the one 
thing that helped. I always had them in 
the house ready for an emergency. I had 
to remember to keep them in the house 
and not run out of them." 

E. Prunes Are "Plebeian" — A Food With- 
out "Prestige" 

"I think of prunes as a good basic food. 
I don't think they are so stimulating; I've 
never longed for one as I do for straw- 
berries or pineapple, for example, except 



82 



UNDERSTANDING THE CONSUMER 



for breakfast. I don't think I would serve 
them to guests. I would serve prune whip 
to guests, but not the plain fruit for des- 
sert. I just don't think of them as a com- 
pany dessert — too plain and substantial. 
I would order them in a restaurant — that's 
different. It would be my own choice." 

"But prunes aren't a festive dish. When 
they are usually mentioned, you think of 
them as being a health dish — something 
that people eat because they need to eat 
them." 

"I have never tried prunes since — at a 
hotel I worked at. We didn't have too 
much food and it wasn't too good — we used 
to have stewed prunes on top of All-Bran. 
Can you imagine that? Sounds horrible! 
We couldn't have this too often. I'm sure 
you understand." 

F. Prunes Are Disparaged as Food for 
"Peculiar" or Old People 

"Constipated and middle-aged people 
would take to eating prunes, because they 
have good reason for it. I can't imagine 
that anybody would eat them for the love 
of it." 

"More females than males, old and 
somewhat dried-up people who tend to be 
health faddists eat prunes. They have cer- 
tain similarities to people who are always 
worried about their health — hypochondri- 
acs. They strain so hard at being natural 
that they aren't natural at all. Young 
people with a normal, healthy organism 
don't need prunes to keep regular." 

"I very definitely connect prunes with 
a certain type of personality. Maybe that's 
why I don't eat them more often. I think 
the people who complain constantly about 
their health and want to be pampered and 
cared for are the people who eat prunes 
regularly. My father, my mother-in-law, 
and my grandmother are people like that. 
There is always something wrong with 
them, and they are forever looking for a 
cure-all. In this way prunes have become 



really more like a medicine than like a 
fruit." 

Some of these unfavorable attitudes 
toward prunes . . . would seem to stand 
in the way of remedial action since they 
are too inconsistent and apparently con- 
tradict the finding that, basically people do 
like prunes. 

Upon deeper anlysis, these attitudes 
appear to be surface reactions which con- 
ceal more basic, deep-seated feelings. It is 
with this deeper layer of motivation that 
we will now concern ourselves. 



The Prune is a "Freak" and "Intruder" 

When we analyzed the most recurrent 
comments and associations given by our re- 
spondents in discussing prunes, we noticed 
that, in a sense, prunes were compared and 
equated with socially undesirable people 
. . . queer, ungiving, self-centered . . . 
people who, in general, are outsiders in 
their own communities and somehow 
"different." 

The more we looked at the record, the 
more strongly we became convinced that 
behind the various superficial negative at- 
titudes toward prunes is the fundamental 
prejudice against the freak, whose ameni- 
ties one enjoys with a bad conscience, 
whose company one shies away from, 
and whose presence one prefers to deny. 

This feeling in relation to prunes can be 
explained as growing out of the feeling that 
nature created fruits for our pleasure and 
gratification, but that prunes deprive us of 
the pleasure to which we are entitled. 
Thus, we feel deprived and frustrated, and 
the resulting reaction is one of aggression. 

Just as a spinster arouses aggression be- 
cause she is believed to deprive the rest 
of her world of her contribution as a 
woman and mother, so it would seem that 
the prune, too, has become a target for 
aggression. 



UNDERSTANDING THE CONSUMER 



83 



The Prune Denies Sensuous Satisfaction 

We have found in earlier food studies 
that food, apart from its nutritional func- 
tion, serves a psychological purpose: it 
helps satisfy those universal needs and de- 
sires which in infancy found gratification 
at the mother's breast. Modern child psy- 
chology is aware of the importance of 
nursing quite apart from its nutritional 
value. For its healthy development, the 
child also needs the pleasure of sucking, 
and of deriving comfort in tactile contact 
with the mother's breast. 

These needs do not cease entirely with 
the end of infancy or with adolescence. 
They continue to exist in a more or less 
unconscious fashion in the grownup indi- 
vidual who seeks socially acceptable ways 
of satisfying them. 

Thus, oral gratification is not confined 
to the health value of food. It includes the 
enjoyment one derives from biting into the 
food, chewing it with the teeth and savor- 
ing it with the tongue and lips, from 
smelling it, and looking at it. 

The prejudice held by many respond- 
ents with regard to prunes is due in many 
cases to the feeling that prunes, in contrast 
to most other fruit, deny the oral gratifica- 
tion which one expects and desires to en- 
joy along with getting the nourishment. 
There are various ways in which prunes are 
felt to frustrate oral pleasure: 

A. The Desire for Lusciousness. Com- 
mon among the terms used to describe 
fruit are "luscious," "rich," "ripe," and 
"juicy." All these words are associated 
with fun and enjoyment. Prunes, how- 
ever, are not described in this manner. 
Rather, respondents visualize prunes as 
hard, wrinkled, dried-out, and unyielding. 

B. The Desire for Potency and Fertility. 
Fruit is a symbol of potency and fertility. 
But when the ripe and full plum is turned 



into a prune, it is felt to be deprived of its 
vital life substance. 

C. The Longing for "Ambrosia." In 
ancient mythology, Ambrosia was the name 
given to the favorite drink of the gods. 
Ambrosia is the perfect liquid food, life- 
giving and entirely satisfying. To the child, 
the mother's milk represents "Ambrosia." 
When he grows up, he forever seeks an 
equivalent to this ideal food, but finds 
only more or less adequate substitutes. 
Ice cream and sweet, juicy and yielding 
fruit come close to the ideal. The prune, 
however, seems to deny the gratification 
of the strong, somewhat infantile desire for 
"Ambrosia." 

D. The Desire for Tactile Gratification. 
As we have stated earlier, tactile gratifica- 
tion is an important component of the en- 
joyment we derive from our food. Our lips, 
tongue, and teeth like to play with and 
feel food-matters of varied textures and 
consistencies. . . . Again, the expectation 
of such pleasures is not borne out with 
regard to prunes. 

E. The Desire for Visual Gratification. 
Fruit is beautiful. It has rich, warm, and 
brilliant colors, that gratify the eye and 
make one's mouth water. Prunes, however, 
appear to evoke different experiences. 

How More Prunes Can Be Sold 

We have outlined the prejudices which, 
on various levels of depth and awareness, 
appear to operate to the detriment of the 
prune and its related products. 

In the last part of this report we shall 
draw our preliminary conclusions as to 
what can be done to get control of these 
irrational forces and, wherever possible, to 
reverse them. 

A. Encouraging People's Hidden Love 
for Prunes. The most important finding so 
far is the fact that most people would like 
to "like" prunes but have been prevented 
from doing so. This means that the over- 



84 



UNDERSTANDING THE CONSUMER 



all promotional trend will have to be in 
the direction of stimulating and encourag- 
ing people to accept their positive feelings 
toward prunes and to free themselves from 
interfering inhibitions. "Everybody loves 
prunes" they should be told, "even if they 
don't say so; and you like them too. Why 
don't you eat what you like?" 

B. Cultural Prejudices Have to Be At- 
tacked. People should be openly con- 
fronted with their prejudices against 
prunes and asked to be fair. Could it be 
that they are wearing, as it were, "dark 
glasses"? They should be shown dramati- 
cally how their prejudice operates. For ex- 
ample, with a pictorial presentation of a 
"white prune" they could be asked: "If 
the prune looked like a white raisin — 
would you still dislike it"? 

C. Be Flexible in a Fast-Changing 
World! People should be helped to realize 
that likes and dislikes are extremely rela- 
tive and changeable, not only from culture 
to culture, but within our culture, from 
fashion to fashion, from year to year. If 
one sticks rigidly to one attitude, one may 
find oneself behind the times. "Be open to 
change and to new suggestions! Your atti- 
tude toward prunes may be a relic and no 
longer tenable." 

D. The Prune— The u Oldest of Modern 
Foods" The fear of the drying and pre- 
serving processes is actually an out-of-date 
fear. Nobody nowadays distrusts frozen 
foods or is afraid of being harmed by 
powdered milk. If the prune were a re- 
cently introduced food, such fears and 
prejudices would have had no chance to 
crystallize. It might be emphasized that 
the prune, in some ways, is the "oldest of 
modern foods. . . ." 

E. Prunes Do Offer Sensuous Satisfac- 
tion and Are Not Just a Food With a 
Purpose. Another promotional trend will 
have to remove from the prune the stigma 
of "utilitarian food." Prunes should be 



eaten for fun; they do offer all sorts of 
satisfactions to the senses. People should 
eat prunes not as "medicine," but because 
they are enjoyable. The feeling of doing 
one's "duty" should be replaced by a sense 
of "let's have fun." 

F. Prunes Are Luscious. The properly 
cooked prune does gratify the desire to sink 
one's teeth into rich, luscious, juicy, "fruit- 
steak." The associations connected with 
full sensuous pleasure can serve to over- 
come the negative ideas of dryness and 
meagerness. 

G. Prunes Give You a Feeling of Po- 
tency and Vitality. Prunes can be made to 
be considered as full and swollen and asso- 
ciated with potency and fertility. "Prunes 
are firm and full and ready to yield their 
rich juice to you." Thus, the notion of 
prune-eaters as weak, impotent, health-con- 
scious people can be combatted in a posi- 
tive way. 

H. Prunes Can Meet the Longing for 
"Ambrosia." Prunes can very well serve to 
gratify the desire of the "grown-up child" 
for "Ambrosia," the heavenly food that is 
sweet, rich, and melts on the tongue. 
As suggested earlier, certain varieties and 
modes of preparation lend themselves par- 
ticularly well to answer the desire for "Am- 
brosia." 

I. Prunes Offer "Tactile" Playfulness. 
People may have to be told "how to eat 
and really enjoy a raw prune." The so- 
called "difficulties" in eating prunes can be 
made into so many playful and gratifying 
activities of all organs of the mouth : suck- 
ing, playing with the pit, and so forth. 
"Prunes are a chewing gum, you can first 
chew and then swallow. . . ." 

J . Making an Asset of a Tactile Frustra- 
tion. The "rough and wrinkled skin" of the 
prune that serves so frequently as a ration- 
alization for all sorts of prejudices can ap- 
pear in a very different light when people 
are reminded that good things often have 




85 



86 



UNDERSTANDING THE CONSUMER 



a rough surface. Perhaps the prune can be 
described as an "interesting" rather than a 
smooth and "pretty" fruit. "Once the 
rough surface is penetrated you will enjoy 
the smooth and meaty inside so much 



more. 



K. Giving the Prune Visual Life and 
Drama. The dark color of the prune, so 
often held against it, has a lot of potential 
qualities of drama and beauty. Strong con- 
trast of yellow and white can be used 
(cream, lemon), and the desire for shini- 
ness and brilliance can be met by the 
jet-like, glittering highlights produced by 
properly prepared and chilled prunes. 

L. Remove From the Prune Its "Ple- 
beian" Connotation. The "prunes are 
prunes" attitude of contempt and indif- 
ference has to be converted into one of 
discernment and respect. The status of 
prunes can be raised to that of raisins by 
advertising their differences and varieties. 
People can be helped to identify the prune 
with more "sophisticated" uses such as 
"prune souffle" and prune whip. They can 
be made conscious of its versatility of 
use alone and in combination with other 
foods. 

M. The Prune Offers a Base for Crea- 
tive Experience. There are many house- 
wives whose reluctance to serve prunes is 



based on the feeling of guilt because they 
are "too easy" to prepare. We found that 
many women like foods that give them 
leeway for their own creative efforts. The 
prune should be presented as an ideal 
"base" for such creative efforts and experi- 
ments. 

On the basis of the information devel- 
oped by the Institute for Research in Mass 
Motivations, the Prune Board's advertis- 
ing agency prepared the proposed adver- 
tisement which appears in Figure 2-J. [The 
original ad was in two colors — bright yel- 
low and black.] 



What light does the report of the In- 
stitute for Research in Mass Motivations 
throw on the problems of the California 
Prune Advisory Board? What difference, 
if any, is there between the actual research 
findings (the factual report of what the 
researchers found, or, in this instance, what 
the interviewees said) and the interpre- 
tation of these findings by the Insti- 
tute? Does this increase or lessen the 
usefulness of the project to the Prune 
Board? Did the Board's advertising agency 
make good use of the report? 



UNDERSTANDING THE CONSUMER 



87 



CASE 2-6 

Prune Research (B) 



In the summer of 1952, the managers 
of the California Prune and Apricot 
Growers Association, a leading marketer 
of prunes, had available some of the pre- 
liminary findings of a motivation research 
study done for the California Prune Ad- 
visory Board, the over-all industry agency, 
by the Institute for Research in Mass Mo- 
tivations, Inc. After consultation with its 
advertising agency, Long Advertising, Inc., 
of San Jose, California, the managers of 
the growers association decided to com- 
mission a separate study. They were inter- 
ested in knowing whether an independent 
survey would confirm the conclusions 
reached in the Prune Advisory Board's 
study. In addition, new data were desired 
on brand and package recognition. 

Long Advertising, Inc., chose the James 
M. Vicary Organization of New York to 
make the new study. It is reproduced be- 
low in abridged form: 

THE PROBLEM 

The purpose of this study is to explore 
the attitudes, ideas and images that con- 
sumers and potential consumers have 
about prunes. 

The semantic approach of this investi- 
gation is designed to discover ideas which 
should be emphasized in advertising copy. 
Such an analysis will also caution the 
advertiser about the ideas that are un- 
favorable and should be avoided or 
circumvented. 

As part of this general survey of atti- 
tudes, we have included a controlled ex- 
periment on the unconscious effect of 



linking the idea of laxative with prunes. 
This experiment demonstrates the mo- 
tivation involved and shows how more 
favorable attitudes can be encouraged. 



METHODS 

In market and opinion research the 
simple, direct question method is most 
often employed. This method very fre- 
quently measures opinion successfully, al- 
though it does not always illustrate or 
reveal basic attitudes upon which action is 
based. Often these attitudes are deeply 
buried and cannot be uncovered by or- 
dinary means. Other techniques had to 
be developed in order to investigate 
deeper, more basic areas of motivation. 
By the use of these more indirect meth- 
ods, findings can be obtained which are 
often lost or ignored in surveys conducted 
along more conventional lines. These 
findings can be of the highest significance 
because they frequently point the way to 
more effective sales promotion. They indi- 
cate underlying attitude areas which are 
favorable and which can be most profit- 
ably exploited. They also detect and ex- 
pose those ideas and attitudes in the pub- 
lic mind which are hostile or uninformed 
and which should be corrected or cir- 
cumvented. 

The word association technique has 
been highly developed in the field of 
clinical psychology. The first experiments 
with stimulus and response words were 
conducted by the English psychologist 
Francis Galton in 1869. Later, the psy- 
chiatrist Carl Jung further developed and 



88 



UNDERSTANDING THE CONSUMER 



popularized the method as a tool for un- 
covering emotional complexes. This, in 
turn, stimulated a vast amount of research 
so that, at present, more is known about 
the word association method than any of 
the other projective techniques. Recent 
studies have also shown that these meth- 
ods can be profitably applied to market- 
ing and public relations problems. 

Single Word Association Test 

In this test the respondent is read a 
list of stimulus words one at a time and 
asked to reply to each with the first word 
that comes to his mind. In this way, and 
by careful planning of the order in which 
the stimulus words appear, the replies be- 
come almost automatic. This test provides 
certain refined semantic measurements 
that will be explained as they are applied 
later in this report. 

Successive Word Association Test 

This technique is somewhat similar to 
the Single Word Association Test discussed 
above. It, however, probes more deeply 
into respondent's attitudes. The respond- 
ent is given a stimulus word and is asked 
to give a series of associated words. Be- 
cause this test extends to deeper levels of 
association, it reveals aspects of connota- 
tion and attitude which are often lost 
with single word associations. The level 
at which a response word occurs can be 
of primary importance in indicating the 
desirability of using the word in adver- 
tising copy. In this study a successive word 
association test of the stimulus word 
prunes was included. The test was de- 
signed to permit a respondent to give as 
many as thirteen responses (word asso- 
ciations). 



Sentence Completions 

This form of investigation is fairly fa- 
miliar and its operation in the interviews 
needs little explanation here. The frag- 
mentary sentences relating to prunes 
which our respondents were asked to com- 
plete were: 



People who don't like prunes 
People who like prunes 



A foil carton for prunes is 



When you give prunes to some chil- 
dren . 



Procedure 

These were the techniques which were 
used to investigate ways of improving ad- 
vertising copy on prunes. The respondents 
were interviewed by experienced interview- 
ers who were specially instructed in the 
use of word association techniques. Two 
hundred interviews were conducted with 
respondents in four cities located in dif- 
ferent parts of the country during the pe- 
riod of June 7, to June 21, 1952. Of these 
interviews, 50 were conducted with men 
in order that we might be able to detect 
any sex-linked differences in attitudes to- 
ward prunes. 

In order to test the importance of back- 
ground differences in our sample, the re- 
spondents were divided into two groups 
— those who had eaten prunes recently 
and those who had not. For practical pur- 
poses, we can assume that most of the 
people who had eaten prunes within the 
week preceding the interview probably eat 
prunes very often. 



UNDERSTANDING THE CONSUMER 



89 



The overwhelming majority of respond- 
ents, 89 per cent of them, said that they 
like prunes when our interviewers asked 
them this direct question. In spite of this 
claim, however, only 44 per cent had 
eaten prunes at some time during the 
seven days previous to the interview. It 
is significant that so many people say they 
like prunes while a much smaller propor- 
tion eat them frequently. A reading of 
the 200 questionnaires shows that people 
say that prunes are "good for you" and 
that they are one of the foods that you 
"ought to eat." 

When was the last time you ate prunes? 



Within the past 3 days 


21% 


4 days to a week ago 


23 


A week to 3 months ago 


33 


3 months to a year ago 


12 


Years ago 


9 


Never ate prunes 


2 


Total (200 cases) 


100% 



PRUNES AS A LAXATIVE 

During the exploratory interviews that 
were conducted in anticipation of the 
present study, it became quite apparent 
that the word prunes often brings to mind 
thoughts of laxatives and their cathartic 
effects. This preliminary finding has been 
confirmed in this study by the results of 
our word association tests which show that 
laxative was a very common association to 
prunes. 

Because of the possibility that such a 
laxative response might have an inhibit- 
ing effect on the number of other asso- 
ciations given to prunes, it was decided 
that only one-half of the questionnaires 
would include laxative as a stimulus word 
on the Single Word Association Test. In 
all other respects, however, the content of 



the questionnaires was to be identical. In- 
terviewers were instructed to administer 
the questionnaires which omitted laxative 
to every other respondent to secure ran- 
domly matched samples. The same num- 
ber of men and women were interviewed 
on each form in all four cities. (For pur- 
poses of simplicity, the questionnaire 
which omits the word laxative will be re- 
ferred to as the "short" form, while the 
questionnaire which includes it will be 
called the "long" form.) 

The procedure is essentially an experi- 
mental one, since comparison of re- 
sponses on the two forms allows us to de- 
tect and measure the effects of the stim- 
ulus laxative on our respondents. Before 
comparing both groups of respondents to 
see what reactions there were to laxative, 
it is necessary to demonstrate that the two 
groups of respondents were essentially 
alike in all other respects. Detailed com- 
parisons as to age, socio-economic status, 
level of education, number of people in 
the household, number of children under 
14 years of age, the last time that they 
ate prunes, and the preferred packaging for 
prunes — all show that the two groups are 
equivalent. 

The only difference between the groups 
is that those people who were interviewed 
on the long form were given the addi- 
tional stimulus word laxative as the fifth 
word of the Single Word Association 
Test. Therefore, since every means of 
checking the similarity of the samples 
shows that they are equivalent, any sig- 
nificant differences in the responses of 
these groups are the result of the intro- 
duction of laxative as a stimulus word. 

Does laxative have the inhibiting effect 
on word associations we expected? It 
should be pointed out that two types of 
inhibition are possible. First, respondents 



90 



UNDERSTANDING THE CONSUMER 



may suppress the number of word asso- 
ciations to prunes, and second, they may 
hold back particular kinds of associations. 

We shall first consider the possible in- 
hibiting effect on the number of associa- 
tions given. On the short form question- 
naire, which did not include laxative on 
the Single Word Association Test, the 
average number of successive word asso- 
ciations to prunes was 5.26, while on the 
long form which did include laxative, 
the average number of responses was 5.56. 
The word laxative, therefore did not in- 
hibit the number of word associations to 
prunes — in fact, it had the effect of 
slightly increasing the number of re- 
sponses. 

As might be expected, the introduction 
of this stimulus word did have an effect 
on the number of times laxative itself 
was given in response to prunes on both 
association tests. As the following table 
shows, there were more than twice as 
many laxative responses on the long form 
as there were on the short one. 



NUMBER OF TIMES LAXATIVE WAS GIVEN 
AS THE ASSOCIATION TO PRUNES 



Single Word Association Test 
Successive Word Association 

Test 
Number of Respondents 


Short* 
Form 

5 

13 
(100) 


Long 
Form 

20 

47 
(100) 



*The short form does not include LAXATIVE as 
a stimulus on the Single Word Association Test. 



significant discovery about the semantic 
relationship between prunes and laxa- 
tive.™ 

Respondents who had eaten prunes re- 
cently did not turn out to be very differ- 
ent from those who had not. About equal 
proportions from each group are of the 
same age levels, belong to families of 
the same size and come from the same so- 
cio-economic classes. Frequent eating of 
prunes is probably due more to deep- 
seated psychological factors than to gen- 
eral background characteristics. This be- 
comes apparent when we compare the 
number of laxative responses to prunes 
given by the recent and nonrecent prune 
eaters on the two forms of the question- 
naire. 

When the same proportion were inter- 
viewed on the short form (where laxative 
is omitted as a stimulus word), approxi- 
mately 30 per cent of both recent 
and nonrecent prune eaters gave laxative 
or a closely related word such as consti- 
pation, as a single or successive associa- 
tion to prunes. On the long form, how- 
ever, where laxative is included as a stim- 
ulus word, the two groups reacted very 
differently. On this form, 35 per cent of 
the nonrecent prune eaters associated the 
idea of laxative to prunes, while 72 per 
cent of those who recently ate prunes 
gave this kind of response. A comparison 
of the two forms for the recent eaters 
shows that the addition of the stimulus 
word "laxative" increased these responses 
by 42 per cent. The corresponding in- 



Normally, such an effect would not be 
considered unusual but would be expected 
on the basis of simple suggestion. How- 
ever, we find that this is not the explana- 
tion. Only certain kinds of respondents 
were responsible for the over-all increase 
in the laxative response to prunes. Further 
analysis shows that this is a clue to a 



16 The remainder of the discussion will deal 
with the interviews of our 150 female re- 
spondents. The findings which follow are 
based upon comparisons of four subgroups 
of respondents; and while there is evidence 
that males follow the same general patterns 
as females, the number of males in each of 
the four subgroups is too small to be able to 
generalize with the same degree of certainty. 



UNDERSTANDING THE CONSUMER 



91 



crease for those who did not eat prunes 
recently was only 4 per cent. 

As can be seen in the tables below, only 
the recent prune eaters were noticeably 
affected by the introduction of the word 
laxative on the long form. 

This reveals an important psychological 
difference between the recent and the 
nonrecent eaters of prunes. Respondents 
who had eaten prunes within the week 
preceding the interview were more sensi- 
tive to the link between prunes and laxa- 
tive. This sensitivity is a latent or con- 
cealed one; for on the short form, where 
laxative was not mentioned by the inter- 
viewer, both groups gave approximately 
the same proportion of laxative responses. 

This suggests that the recent prune 



eaters were "holding out" on the inter- 
viewers when the short form was used. 
Further analysis of the successive re- 
sponses on the short form reveals that the 
recent group gave associations which were 
socially acceptable substitutes or symbols 
for the idea of laxative. When respond- 
ing to prunes on the short form, 33 per 
cent of the recents gave a euphemistic re- 
sponse such as healthy, good for you or 
keeps you fit. When the long form was 
used, only 20 per cent of these respond- 
ents gave euphemistic associations. On 
the other hand, for people who did not 
eat prunes recently, there was no signifi- 
cant decrease in the use of euphemisms 
on the long form. 

This clearly shows that the idea of lax- 



PROPORTION OF RESPONDENTS WHO ASSOCIATED THE IDEA OF 
LAXATIVE TO PRUNES 





Recently 


Ate Prunes 


Nonrecents 




Short* 


Long 


Short* Long 




Form 


Form 


Form Form 


Total Number of Respondents 


(33) 


(35) 


(42) (40) 


Proportion Who Associated the 








Idea of Laxative to PRUNES 


30% 


72% 


31% 35% 


Effect of adding LAXATIVE as 








a Stimulus 




h42% 


+4% 



*The short form does not include LAXATIVE as a stimulus on the Single Word 
Association Test. 



PROPORTION OF RESPONDENTS WHO GAVE A EUPHEMISTIC 
RESPONSE FOR LAXATIVE ON SUCCESSIVE ASSOCIATIONS 
TO PRUNES 





Recently 


Ate Prunes 


Nonrecents 




Short* 


Long 


Short* Long 




Form 


Form 


Form Form 


Total Number of Respondents 


(33) 


(35) 


(42) (40) 


Proportion Giving a Euphemistic 








Response 


35% 


20% 


21% 18% 


Effect of Adding Laxative as a 








Stimulus 




•13% 


-3% 



*The short form does not include LAXATIVE as a stimulus on the Single Word 
Association Test. 



92 



UNDERSTANDING THE CONSUMER 



ative has a special meaning to women 
who eat prunes frequently. It follows 
from the data that these women have 
had a constipation problem about which 
they are highly sensitive. Further analysis 
shows that hearing the word "laxative" ac- 
tually reduces the general level of anxiety 
for these respondents. This state of relief 
produced by the symbol laxative had meas- 
urable effects which pervaded the entire 
interview. The remaining sections of this 
report will fully demonstrate this con- 
clusion and will discuss its implications 
for advertising copy. 

In a word association test, one of the 
most reliable signs of anxiety is a delay of 
over three seconds before giving a re- 
sponse to a stimulus word. Extensive clin- 
ical and experimental work with the word 
association technique has demonstrated 
that a long reaction time or complete 
blocking is generally the result of an emo- 
tional disturbance associated with the 
stimulus. 

When interviewed on the short-form 
questionnaire, our respondents partially 
or completely blocked on about 25 per 
cent of their responses to stimulus words 
following laxative in the Single Word As- 
sociation Test. Both respondents who re- 
cently ate prunes and the nonrecents had 
about the same level of anxiety on the 
short form. On the long form, however, 



where laxative appears as the fifth stim- 
ulus word, the recent prune eaters blocked 
on 8.1 per cent fewer of their responses. 
The anxiety level of the nonrecents did 
not show any significant change at all. 

It is clear that the addition of the stim- 
ulus word laxative affected the responses 
that recent prune eaters gave to the other 
stimulus words. This indicates a general 
reduction in anxiety level. Once again we 
see that the nonrecent prune eaters do 
not give any indication of having been 
affected by the addition of the stimulus 
word laxative. 

In the field of clinical psychology, it is 
found that anxieties are often expressed 
as hostility toward others — especially peo- 
ple not like oneself. When we examine 
the responses to the sentence comple- 
tions, we find that the recent prune eat- 
ers generally made derogatory or uncom- 
plimentary remarks about people who 
don't like prunes. They said that these 
people are ''grouchy;" "fussy," "consti- 
pated," "silly," "don't know what's good 
for them," "aren't healthy," and so forth. 
These expressions of hostility toward peo- 
ple who are not like themselves were 
made by 70 per cent of the recents when 
the short form was used. (Researchers in 
free association techniques have found that 
a refusal to answer indicates some kind of 
a negative attitude.) When the question- 



BLOCKED RESPONSES FOR THE 25 WORDS FOLLOWING LAXATIVE 
ON THE SINGLE WORD ASSOCIATION TEST 





Recently Ate Prunes 


Nonrecents 




Short* Long 


Short* Long 




Form Form 


Form Form 


Number of Respondents 


(33) (35) 


(42) (40) 


Total Number of Responses 


825 875 


1,050 1,000 


Number of Blocked Responses 


212 163 


261 247 


Proportion of Blocked Responses 


26.7% 18.6% 


24.8% 24.7% 


Effect of Adding LAXATIVE as 






a Stimulus 


-8.1% 


-0.1 



*The short form does not include LAXATIVE as a stimulus on the Single Word 
Association Test. 



UNDERSTANDING THE CONSUMER 



93 



naire with the word laxative was used, 10 
per cent fewer of the recents showed hos- 
tility toward people who don't like prunes. 
The mention of the word laxative in the 
word lists not only reduced anxiety, but 
also reduced the hostility of recent prune 
eaters toward people not like themselves 
in this respect. 



ATTITUDES OF RECENT PRUNE EATERS 

TOWARD PEOPLE WHO DON'T LIKE 

PRUNES 

Short* Long 
Form Form 



(33) 



(35) 



40% 



Number of Respondents 
Proportion Giving Favorable 

or Neutral Responses 30% 

Proportion Giving Unfavorable 

Responses or Not 

Answering 70% 60% 

100% 100% 

"The short form does not include LAXATIVE as 
a stimulus on the Single Word Association Test. 



The majority of respondents who did 
not recently eat prunes gave favorable or 
neutral responses when they completed 
the sentence: People who like prunes 

. This again demonstrates the 

high sanction given to prunes at the con- 
scious level. The word laxative produced 
a slight drop in hostility for the nonre- 
cents. Although only 17 per cent of them 
were unfavorable toward people who like 
prunes, this proportion became 12 per 
cent when laxative was added to the word 
list much earlier in the questionnaire. For 
both groups, then, laxative had the effect 
of increasing favorable feeling toward peo- 
ple who are different. It became easier 
to identify oneself with another kind of 
person when anxiety was reduced. 

Earlier we saw that, when laxative was 
included in the word list, it induced more 
respondents to give a laxative association 
to prunes. The following table shows that 
when the respondents were induced to 
give this kind of a response, it resulted in 



an increase in their favorable opinions to- 
ward prunes. 

DO YOU LIKE PRUNES? 

Respondents Who Did Not Associate the Idea 
of Laxative With Prunes 



Respondents 

Proportion Saying They Like 
Prunes 



Short* 
Form 

(52) 
77% 



Long 
Form 

(36) 
75% 



Respondents Who Did Associate the Idea of 
Laxative With Prunes 



Respondents 

Proportion Saying They Like 
Prunes 



Short* Long 

Form Form 

(23) (39) 

78% 90% 



♦The short form does not include LAXATIVE as 
a stimulus on the Single Word Association Test. 

We have seen that women who re- 
cently ate prunes are sensitive about the 
problem of constipation. When the word 
laxative was not mentioned by the inter- 
viewer, direct expression was inhibited in 
this area. At the same time, these re- 
spondents revealed their sensitivity by 
giving euphemistic associations. When 
laxative was mentioned, the reduction of 
anxiety permitted more direct expression 
of the idea that prunes have cathartic ef- 
fects. There was also a lessening of hos- 
tility toward people who do not like 
prunes and an increase in favorable opin- 
ion of prunes at the conscious level. The 
pattern we have found in this study holds 
together at every level of mental activity 
tested. 

The data indicate that, to properly ex- 
ploit the core of the present market, ad- 
vertising copy should mention the laxative 
feature of prunes. Euphemisms such as 
healthful should be avoided since they 
do not provide the sanction required to 
remove the emotional block in this area. 
No evidence was found that people who 



94 



UNDERSTANDING THE CONSUMER 



did not eat prunes recently would react 
unfavorably if the laxative effects of 
prunes were suggested. 

Other experiments have shown that 
people who have disturbed responses on 
individual words cannot recall correctly 
material relating to these subjects. Clini- 
cal experience has found that such dis- 
turbances have strong motivational impli- 
cations and that normal behavior patterns 
are distorted. To allow tension and un- 
expressed concern over the laxative effects 
of prunes to remain unconscious destroys 
many of the possibilities of getting other 
messages across to consumers. 



WORD ASSOCIATIONS TO PRUNES 

The Single Word Association Test per- 
tains to the responses which are first 
brought to mind for a given stimulus 
word. Since it was desirable to get at the 
deeper levels of association, respondents 
were also given a Successive Word Asso- 
ciation Test for prunes. A comparison of 
the most common responses of both tests 
indicates the words that are especially im- 
portant at the deeper levels. 

The most frequent single associations 
to prunes were fruit, plums, eat, apricots, 
juice, laxative, and breakfast. An inspection 
of the top successive responses shows that 
eat, apricots, and breakfast do not stay at 
the top of the list but are replaced by 
dried, stewed, and healthy, which take on 
significance at the deeper associational 
levels. Experience has shown that re- 
sponses which are frequent only on suc- 
cessive associations often move to the 
first-response level over a period of time. 
When this happens, it reflects the fact 
that the stimulus word has developed 
new major connotations. An important 
feature of word association studies is that 
they identify these words of potential 
growth. With this knowledge, the adver- 



tiser can build up the words which are 
favorable for his purposes, and de-empha- 
size the words that have undesirable con- 
notations. The associations to prunes 
which show particularly high growth po- 
tential are dried and stewed. These are 
two top successive associations, although 
neither appears as a frequent single word 
association response. 

TOP ASSOCIATIONS TO PRUNES* 



The Underlined Words do not Appear as Top 
Responses on the Opposite List 



Single Word 








Association 




Successive Word 




Test 




Association Test 




Fruit(s) 


20 


Dried 


24 


Plum(s) 


7 


Stewed 


21 


Eat 


6 


Fruit(s) 


20 


Apricot(s) 


5 


Laxative(s) 


20 


Juice(s) 


5 


Plum(s) 


19 


Laxative(s) 


5 


Juice(s) 


16 


Breakfast 


4 


Healthy 


14 


Dessert(s) 


4 


Dessert(s) 


12 



♦Only the responses of the 100 respondents who 
were interviewed on the short form are included in 
this table. 



We have already seen that words such 
as healthy are used as euphemisms for the 
laxative effects of prunes. In the preceding 
chapter, it was recommended that these 
euphemistic words be strictly avoided in 
advertising copy because they do not 
break through the anxiety related to the 
laxative area. 

Fruit, of course, is a favorable response 
and, like plums, is a strong association at 
both the immediate and deeper levels. 
But what about the word dried, which is 
the most frequent response and may 
have a high potential for growth? Of the 
30 stimulus words on the Single Word 
Association Test, dried had the greatest 
proportion of blocked responses. This 
blocking reaction by over one-half of the 
respondents indicates that dried is a highly 



UNDERSTANDING THE CONSUMER 



95 



PROPORTION OF RESPONDENTS WHO 
BLOCKED TO THE STIMULUS WORDS ON 
THE SINGLE WORD ASSOCIATION TEST* 

Proportion Giving 
Stimulus Word Blocked Responses 

Number of Respondents (100) 

Dried 53% 

Foil Carton 50 

Dried Fruit 36 

Noise 33 

Buyer 32 

Prunes 29 

Money 29 

Advertising 26 

Stewed Fruit 24 

Coupons 24 

Accident 23 

Healthy 22 

Tasty 20 

Housework 18 

Gift 18 

Looks Appetizing 17 

Teach , 17 

Reader's Digest 16 

Good Flavor 16 

Family 16 

Story 15 

Desserts 14 

Wrinkled 13 

Frozen Food 12 

Sticky 10 

Wash Day 10 

Garden 8 

Comb 7 

Clock 4 

♦Only the 100 respondents who were interviewed 
on the short form are included in this table. 



disturbing word. Dried fruit was also one 
of the most disturbing stimulus words on 
the list. 

The table to the left gives the propor- 
tion of blocked responses for the 30 sin- 
gle word association stimulus words. 

When we questioned our respondents 
about packaging for prunes, it became 
clear that they wanted their prunes to be 
moist, tender, and fresh. Because of this 
and the highly disturbed response reac- 
tions to dried and dried fruit, we strongly 
recommend that the "dried-up" charac- 
teristic of prunes be played-down in all 
advertising copy and illustrations. 

The advertiser may take advantage of 
the fact that fruit and plums are frequent 
responses at both levels. Giving more em- 
phasis to these already common and fa- 
vorable associations would tend to coun- 
teract the strong unfavorable image of 
prunes as dried. 



Are these research findings at odds 
with those of Case 2-5? What further or 
different light do they throw on consu- 
mers' attitudes toward prunes? What ad- 
vertising appeals should result from these 
findings? From these findings when coupled 
with those of Case 2-5? 



96 



UNDERSTANDING THE CONSUMER 



In 1955, Dow's manager of organic 
chemical sales asked the manager of 
market research to investigate long-range 
trends in the consumption of ethylene 
glycol. This information was needed to 
help the company's top management de- 
cide whether Dow's capacity to produce 
ethylene glycol should be increased. 

The report of the market research de- 
partment's organic chemicals specialist fol- 
lows herewith: 



PURPOSE 

The purpose of the present report is 
to prepare a preliminary long-range pro- 
jection of ethylene glycol requirements 
for automotive antifreeze. Most of the 
data utilized in the preparation of this 
present report have been obtained from 
the technical literature, and are subject to 
considerable uncertainty. The long-range 
nature of the projection of motor vehicle 
registration and production is predicated 
upon continuance of a normal economy 
and the absence of any national emer- 
gencies. 

From the reports in the literature, an 
attempt has been made to evaluate the 
status of automotive gas turbine develop- 
ment and to project the impact that gas 
turbine introduction would have on anti- 
freeze requirements. 

A secondary purpose of this report is to 
point out the nature of the data available 
at present. By recognizing the inadequa- 
cies of the data, it is our intention to carry 
out field contacting to confirm (or to 
modify) the underlying assumptions em- 
ployed in this report. 



CASE 2-7 

Dow Chemical Company (A) 



PRODUCERS AND CAPACITIES 

At the present time there are seven 
producers of ethylene glycol. Plants are 
operated in 11 locations and have an an- 
nual capacity of about 930,000,000 pounds 
per year. 



PRODUCTION FIGURES 

Annual production figures for ethylene 
glycol are published by the United States 
Tariff Commission. Figure 2-M shows pro- 
duction figures for the period 1930-1954; 
also shown for some years, where data 
are available, are sales figures. These data 
are tabulated in Table 2-K. 

Table 2-L lists ethylene glycol produc- 
tion figures for the years 1947 through 
1955 by months; these data are shown 
graphically in Figure 2-N. 

From 2-M and 2-N we see that the 
period 1947-1951 was characterized by a 
rather regular increase in ethylene glycol 
production, which averaged an increase of 
about 90,000,000 pounds per year. In 
1952, however, production jumped to 
761,000,000 pounds, or an increase of 
about 162,000,000 pounds over the 1951 
figure. In 1953, production early in the 
year started off at the same rate as in 
1952; by March, however, production 
decreased very substantially with the re- 
sult that 1953 production was 630,000,000 
pounds. In 1954, production again de- 
creased, to a level of about 600,000,000 
pounds. 

The first four months of 1955 showed 
glycol production at a rather moderate 



UNDERSTANDING THE CONSUMER 



97 



level, approaching that of 1954. During 
May and June, however, production in- 
creased sharply, giving for each of these 
months production figures substantially in 
excess of any previously recorded figures. 
Assuming a production capacity of 880,- 
000,000 pounds annually, it would appear 
that during May and June the glycol plants 
were operating at close to capacity. 

By using the first six months' figures 
(Table 2-L) for the past eight years, it 
is possible to determine what percentage 
of annual production is, on the average, 
produced during the first six months. 
Applying this average percentage figure 
(47.66 per cent) to 1955 production, it is 
estimated that 1955 production will be 
about 780,000,000 pounds. 



TABLE 2-K. PRODUCTION AND SALES OF 
ETHYLENE GLYCOL 



(Thousands of Pounds) 









Sales as % 


Year 


Production 


Sales 


of Production 


1930 


18,000 


-- 


-- 


1931 


16,000 


-- 


-- 


1932 


11,000 


-- 


-- 


1933 


13,000 


-- 


-- 


1934 


42,000 


-- 


-- 


1935 


66,000 


-- 


-- 


1936 


82,000 


-- 


-- 


1937 


102,000 


-- 


-- 


1938 


140,000 


-- 


-- 


1939 


120,000 


-- 


-- 


1940 


133,000 


-- 


-- 


1941 


151,543 


-- 


-- 


1942 


195,000 


-- 


-- 


1943 


186,834 


-- 


-- 


1944 


202,451 


-- 


-- 


1945 


205,087 


-- 


-- 


1946 


192,176 


-- 


-- 


1947 


266,679 


-- 


-- 


1948 


366,712 


265,262 


72.4 


1949 


439,746 


-- 


-- 


1950 


519,013 


398,186 


76.7 


1951 


596,737 


445,713 


74.7 


1952 


760,959 


599,025 


78.7 


1953 


624,324 


224,522 


35.9 


1954 


600,159 


-- 


-- 


1955 


780,000(est.) -- 


-- 



If one extrapolates the ethylene glycol 
production curve (Figure 2-M) for the 
period 1946-1951, it appears that the 
overproduction of 1952 should have been 
consumed during the period 1953-1954; 
that is, the area above the extrapolated 
curve appears to equal the area below the 
curve. Furthermore, the estimated figure 
for 1955 falls on the extrapolated curve. 
Actually, it is felt that the extrapolated 
curve as drawn is rather conservative since 
it predicts an annual production increase 
of only 5.5 per cent; by contrast, the 
average growth for 1946-1951 was in ex- 
cess of 20 per cent. 



END-USE PATTERN OF ETHYLENE GLYCOL 

Ethylene glycol is used for a variety of 
applications, with the largest single use 
being for automotive antifreeze. Table 
2-0 presents the end-use pattern for eth- 
ylene glycol in 1950 as estimated by the 
Association of American Soap and Glyc- 
erine Producers. . . . 

Although this report is concerned pri- 
marily with the long-range projection of 
ethylene glycol for antifreeze require- 
ments, a knowledge of nonantifreeze uses 
is required to bring the over-all picture 
into focus. It is, therefore, recommended 
that Dow's past glycol sales be scrutinized 
to ascertain whether definite trends are 
indicated. As another part of this picture, 
an attempt should be made to determine 
the national end-use pattern. 



THE ANTIFREEZE PICTURE 

To project the future requirements for 
ethylene glycol for automotive antifreeze 
requires that we first estimate the future 
motor vehicle registrations and, secondly, 
that we consider the past pattern of anti- 
freeze use. 



m 
in 


o 
c- 
co 


CO 


CO 


o 

CO 

in 


00 


CO 
CM 


oo 
m 

CO 


^ 


CM 

CO 


CO 

o 


CO 

in 

Ol 


CO 
CO 


t- 


■^ 


in 


in 


CO 


CO 


en 


o 

CM 


00 


i-H 
CM 


CM 
CM 


CM 


CM 
CM 



00 

o 
o 


CM 

CO 

o 


00 

m 


CO 
CO 




CM 


o 


CO 
OS 


CO 
CO 


oo 

CM 
CO 


CM 

00 


o 
t- 

co 


00 
CM 


CO 
CM 


CO 
CM 


00 
CM 


oo 

CM 


CM 

CO 


o 

CO 


o 

CO 


CO 
CO 


CM 

CO 


CO 
CO 


l— 1 
CO 


CO 


CO 
CO 



CO 


,_, 


i-t 


en 


o 


in 


00 


CO 


IT- 


CM 


en 


,-H 


■<* 


■^ 


o 


CM 




00 


CO 




OS 




CO 


Oi 


in 


Oi 


•«* 


CM 


en 


O 


CM 


o 


CO 


to 


m 


o 


CO 


m 


CO 


O 


m 


in 






























CO 


in 


00 


CO 


CO 


c- 


00 


CO 


CO 


1— 1 


CO 


00 


O 


C- 


CO 


CM 


CO 


CO 


CO 


co 


CO 


CO 


CO 


CO 


CO 


CO 


CM 


o 

CM 



co 


o 


CO 


CO 

in 


CM 

o 


o 

CM 

00 


CO 


CO 
00 
CM 


00 

00 


in 

00 
CO 


oo 
O 
oo 


CM 

Ol 
CD 


CM 
CM 

in 


00 

CO 


05 
CO 


CM 


Oi 
CO 


CO 
CO 


Oi 
CO 


CO 


CO 


•<* 


in 


Ol 


o 
m 


o 
in 



on 


Ol 


CO 


in 


■^ 


"<*< 


rH 


00 


CM 


CO 


o 


00 


tr- 


00 


in 


o 


05 


oi 


CO 


00 


ee 


m 


a> 


CO 


















r- 


Ol 


CO 


in 


■^ 


in 


oo 


,-H 


TT 


■^ 


m 


in 


in 


in 


oi 
in 


00 
CM 



in 


CM 


CM 


in 


■<*« 


CD 


r-l 


tr- 


in 


en 


t- 


Ol 


CO 


CO 


m 


co 


o 


Ol 


co 


oo 


CO 


Tj< 


00 


•"* 


















in 


c- 


00 




c- 


CO 


o 


CO 


CO 


CO 


m 


CO 


in 


CO 


CO 


00 
CO 



CO 


D- 


00 


00 


oi 


H 


CO 


co 


Tf 


Ol 






o 


CM 


c- 


00 


m 


"«* 














•^ 


m 


o 


CO 


CO 


CO 


CO 


in 


CD 


in 


in 


in 



H 


00 


CO 


CO 


CO 


,-H 


Ol 


^ 


CO 


c- 


CO 


oo 


co 


rr 


in 


"tf 


co 


Ol 


CO 


CM 


oo 
















Ol 


in 


in 


■^ 


CO 


o 


CO 


■* 


-<* 


■^ 


■^ 


"<* 


CO 

co 


CO 



p en 


«* 


CD 


CO 


00 


in 


,_| 


*tf 


CO 


m 


CM 


os 


01 


o o> 


CO 


CO 




CD 


CD 


T-\ 


00 


in 


Ol 


CO 


m 




■J 00 


Ol 


"«r 


CM 


CO 


en 


"^ 


CM 


CO 


tj< 


o 




CO 


•> en 


,-H 


m 


00 


■«* 


T-t 


CD 


rH 


O) 


CO 


•*!< 


Ol 


Oi 


i CO 


Tj< 


■^ 


•^ 


^ 


in 


m 


in 


m 


m 


in 


Ol 

in 


CD 
CM 





Ol 


co 


m 


CD 


■"^ 


on 




en 


o 


O 


in 


in 


o 




in 


CM 


CO 


t- 


00 


en 


















r-H 


in 


O 


in 


CM 


■f 


' -l 


in 


in 


CO 


in 


ir- 


c- 



u 

o * 

fe S < 



>> 
o a 



98 



1000 






















ETHYLENE GLYCOL 
U.S. PRODUCTION 8 SALES 




1 


800 






I _- 






A. i*ti r 








vjw^HH 


600 


























r i i 




400 








/\/' \ -SALES 








/ / 
/ • 


\ 
\ 

\ 




200 










/ / 
/ / 


\ 
\ 
\ 
\ 




















MILLIONS 
OF POUNDS 




'—PRODUCTION 


100 






















80 


























60 
























40 




























20 

10 



























1930 



1940 1950 

Figure 2-M 



I960 

























80 




tIMTLtNt (9LTCUL 

U.S. MONTHLY PRODUCTION 












w 
















/ 


70 










v — W" 








— 


60 


















PRODUCTION 
RATE, M LBS/MO. 


/ — 












50 
40 
















— 


— hi 














— 


30 
20 














— 


1948 


1949 


1950 


1951 


1952 


1953 


1954 


1955 



Figure 2-N 



100 



UNDERSTANDING THE CONSUMER 



TABLE 2-0. END-USE PATTERN OF 
ETHYLENE GLYCOL- 1950 





A.A.S. &G.P. 


Estimate 




of National 


Pattern 


End- Use 


MM Lbs. 


% 


Antifreeze 


381.4 


76.3 


Dynamite 


25.0 


5.0 


Cellophane 






Printing Inks 


21.5 


4.3 


Textile Chemicals 






Hydraulic Fluids 


-- 


-- 


Alkyd Resins 


7.0 


1.4 


Adhesives 


1.5 


0.3 


Miscellaneous 


63.6 


12.7 



TOTAL 



500.0 



100.0 



A. Projected Motor Vehicle Population 

Figure 2-P shows the motor vehicle 
registrations for the United States for 
the period 1925-1953. In Figure 2-Q, data 
are presented which show the production 
of motor vehicles in the United States for 
the period 1925-1954. For 1955, it is es- 
timated that the total production of mo- 
tor vehicles is 7,170,000 units, of which 
5,800,000 units are passenger cars, and 
1,370,000 units trucks and buses. 

An examination of the data for registra- 
tion and production should permit pro- 
jections of total vehicle registrations for 
the future. As indicated by the data pre- 
sented in Figure 2-Q, the production of 
both passenger cars and trucks is subject 
to considerable fluctuation. It is assumed 
here that, on the average, the production 
of passenger cars will increase by 50,000 
units annually, while bus and truck pro- 
duction will increase by 40,000 units an- 
nually. Although the fluctuations in the 
production curve are quite large, an esti- 
mate of the average annual increase can 
be obtained by drawing the best straight 
line; using this procedure for the period 
1925-1952, the average annual increase 
in passenger car production is found to 



be about 50,000-70,000 units annually 
while that for truck and bus production 
is about 40,000. On this basis, the figures 
that have been selected for the present 
projection are of the proper order of 
magnitude. 

An examination of the production and 
registration data permits an estimate of 
the rate at which motor vehicles are made 
obsolete. Again, the fluctuations in the 
production curve make it difficult to ob- 
tain any firm figure on obsolescence; 
however, it appears that about three ve- 
hicles are made obsolete for every four 
new vehicles produced. In other words, 
each year the registration will increase by 
an amount equal to 25 per cent of the 
new motor vehicles produced that year. 

Using these assumed figures for produc- 
tion and rate of obsolescence, the motor 
vehicle registrations can be projected, as 
shown in Figure 2-P. It is indicated that 
by 1965 there will be about 78,500,000 
motor vehicles registered and that by 
1980 this figure will have increased to 
111,000,000. These figures are in sub- 
stantial agreement with projections by 
people in the automotive industry who 
foresee 81,000,000 vehicles in 1965 and 
in excess of 100,000,000 vehicles by 1980. 
Table 2-R lists our estimated figures for 
motor vehicle population through 1970. 



B. Past Pattern of Antifreeze Use 

Data relating to the consumption of 
antifreeze are rather meager and subject 
to some uncertainty. The Chemical Eco- 
nomics Handbook, published by the Stan- 
ford Research Institute, has compiled an- 
tifreeze data based on weighed selections 
from available data in the literature. How- 
ever, when one refers to the references 
cited, it is found that the antifreeze con- 
sumption data were, in turn, estimated 
from production figures. Since these are 



UNDERSTANDING THE CONSUMER 



101 



1925 



100 




















U.S. MOTOR VEHICLE 
REGISTRATIONS 












80 
























._ _-— """"" 














' - -- " — " 




60 










TOTAL 


^-^ 1 




40 










^x^L- ^^PASSENGER CARS 




















20 


^^^^ 


^ -""" 


^ ^ 




^ 














































MILLIONS OF | 
VEHICLES | 




- 
















— «— 






8 
























•-^TRUCKS Q BUSES 




s 








































































2 



















1935 



1945 
Figure 2-P 



1955 



1965 







U.S. MOTOR VEHICLE 








6-0 




PRODUCTION 




















~\7 


4.0 












V \ 




\— /- 








2.0 






/ PASSEN 


GER CARS 














MILLIONS OF | 
VEHICLES i 




1 






















08 
























0.6 






















04 


















V /trucks a BUSE 


s 






02 















1925 



1935 



1945 



1955 



Figure 2-Q 



102 



UNDERSTANDING THE CONSUMER 



the only data available, they will be used 
for the present projection. 

Table 2-S presents antifreeze data for 
the period 1934-1952, abstracted from 
the aforementioned source. Also shown 
are the percentage figures for the portions 
of the total antifreeze market enjoyed by 
methanol and ethylene glycol. Figure 2-T 
shows the percentage data in graphical 
form. 

Although not shown in the table, ethyl 
alcohol, which was the major antifreeze 
material in 1934, has practically left the 
market as an antifreeze. The disappear- 
ance of ethanol as an antifreeze has led to 
an increasing use of methanol and eth- 
ylene glycol as antifreezes. Propylene 
glycol is used to a limited extent but is 
not considered a serious factor, since one 
cannot determine the antifreeze concen- 
tration by a simple hydrometer test. 

As seen from Figure 2-T, ethylene 
glycol has cornered an increasing share 
of the antifreeze market, especially dur- 
ing the period from 1947 to the present. 
Extrapolation of the trend line would in- 
dicate that at the present time ethylene 
glycol should account for 55 per cent of 
the total antifreeze market. In an article 
in Chemical and Engineering News (No- 
vember 15, 1954), DuPont people are 
quoted as saying that glycol accounts for 
53 per cent of the antifreeze market; in 
the same article Carbide and Carbon rep- 
resentatives claim the glycol share of the 
market is 55 per cent. 

There is no question but what the con- 
sumption of antifreeze will increase, due 
to increased car registrations. The impor- 
tant point is — what is the level of anti- 
freeze consumption per registered vehi- 
cle? By using the data of Table 2-S, along 
with the figures for motor vehicle regis- 
tration, a trend line can be constructed 
for the period 1935-1952. Such a graph is 
shown in Figure 2-U. It is seen that the 



antifreeze consumption, in terms of gal- 
lons per vehicle registered (which includes 
passenger cars, trucks and buses, and pub- 
licly owned vehicles but does not include 
tractors and other off-the-road equip- 
ment) has been, roughly, between 1.5 
and 2.0 gallons per vehicle. However, the 
curve would appear to be trending up- 
ward as evidenced by the higher peaks 
and lower valleys of the curve. Although 
of limited utility, the Gompertz equation 
has been applied to this curve and the 
result indicates a leveling off at 1.92 gal- 
lons per registered vehicle. 17 That this fig- 
ure is of the right order of magnitude is 
confirmed by the general feeling in the 
antifreeze industry that the per vehicle 
consumption of antifreeze is about two 
gallons; the Jefferson Chemical Company 
has estimated 1.8 gallons of antifreeze 
consumed per registered vehicle. 



C. Preliminary Estimate of Future Antifreeze 
Consumption 

The projections of motor vehicle popu- 
lation and the data on past antifreeze con- 
sumption have been used to estimate fu- 
ture antifreeze requirements. Table 2-V 
shows projected ethylene glycol require- 
ments for antifreeze predicated on the 
following assumptions (Case A) : 

i. The motor vehicle registrations will 
be as shown in Table 2-V. 



17 This equation expresses the proportions of 
a so-called Gompertz curve, which starts 
almost horizontally, then accelerates until 
it is almost vertical, 
then decelerates at 
the same rate, until 
it finally levels off 
again. This has 
been the life- 
pattern of demand 
in many industries. 



UNDERSTANDING THE CONSUMER 



103 



ii. Antifreeze consumption per vehicle 

is assumed to be 1.92 gallons per 

year, 
iii. Ethylene glycol will constitute 55 

per cent of the total antifreeze 

market. 
Another basis for estimating ethylene 
glycol antifreeze requirements is to as- 
sume a static per car antifreeze consump- 
tion, but assume a gradual increase in the 
portion of the total antifreeze market en- 
joyed by ethylene glycol. Case B, which 
assumes that the ethylene glycol portion 
of the antifreeze market will gradually in- 
crease from 55 per cent in 1955 to 65 per 
cent in 1965 and then remain at the lat- 
ter level, is set forth in Table 2-W. Since 
the percentage of the antifreeze market 
enjoyed by glycol has been trending up- 
ward, Case B might seem more reason- 
able than the assumptions incorporated 
in Case A. Figure 2-X shows graphically 
the projected annual United States re- 
quirements for ethylene glycol for anti- 
freeze use for the assumed Case A and 
Case B. 

These figures cannot be considered as 
any more than rough estimates since we 
have not considered such factors as (1) 
shifts in population, or (2) change in 
the ratio of trucks to passenger cars. Fur- 
thermore, tractors and other off-road ve- 
hicles and engines have not been included 
in the present projection. 



D. The Gas Turbine Question 

All of the estimates presented thus far 
have been predicated on the assumption 
that piston engines will constitute the 
whole of the automobile market through 
1970. The possible advent of the auto- 
motive gas turbine would have a tremen- 
dous impact on the requirements for au- 
tomotive antifreeze since gas turbine cars 



would not have radiators and would not 
require antifreeze. 

That the gas turbine will be used for 
automotive purposes appears to be an 
eventuality accepted by a majority of key 
people in the automotive and petroleum 
industries. The big question mark is — 
how and when will automotive gas tur- 
bines be introduced? There is much spec- 
ulation in the technical literature as to 
the time when gas turbines will be a ma- 
jor factor in the automotive market and 
guesses range from 15 years to 25 years 
to achieve a 100 per cent transition. 

National Petroleum News' automotive 
editor, Holger Ridder, recently completed 
a six-month study of the progress and 
present status of automotive gas turbines. 
The study included interviews with 18 
top automotive and oil industry engineers 
and research directors. Confidential oil 
company research reports were examined 
and checked with automotive gas turbine 
men for accuracy. 

Of the automotive and oil industry men 
interviewed, more than 80 per cent be- 
lieve passenger cars — certainly trucks and 
buses — will be powered by gas turbines 
within 10 years. More than half predicted 
that gas turbines will appear first in pas- 
senger cars in five years and eventually 
will replace today's high-compression pis- 
ton engine. 

Oil company opinion varies widely on 
the how and when of gas turbines. Of 14 
oil companies interviewed, eight see the 
gas turbine in action by 1970. Four oth- 
ers thought 1975, one had no opinion, 
and the other said "never." 

The predictions of the oil companies' 
for large-scale turbine production is as 
follows: one company expects it by 1975, 
five believe it will come by 1970, with 
the remainder of the opinion that it is a 
long way off. 

Chrysler and Ford, who seem to have 



104 



UNDERSTANDING THE CONSUMER 



made the most progress on automotive 
gas turbines, are firm in the belief that 
the gas turbine for passenger cars is in- 
evitable and possible within the next 15 
years. General Motors, on the other hand, 
sees the first practical application of tur- 
bines in trucks and buses, and that, some 
time off. 

A 100 per cent transition from pistons 
to turbines probably would take 25 years. 
Even the most optimistic estimate places 
it 20 years away. 

There is agreement in the automotive 
circles that a turbine, once proven prac- 
tical and acceptable to the public, will 
signal a rush of all car and truck manu- 
facturers to build turbines. Assume tur- 
bines first appear in specialty cars by 
1965; there will follow at least two years 
of testing the turbine market and ironing 
out production problems. 

Once the pilot stage has been com- 
pleted, about three years must be allowed 
for tooling and building plant facilities for 
the first mass-production turbines. That 
brings us to 1970. 



CONCLUSIONS 

A. Present ethylene glycol capacity in the 
United States is about 930,000,000 
pounds annually. An announced ex- 
pansion project brings this figure to 
1,020,000,000 pounds within two 
years. 

B. The major use of ethylene glycol is 
for automotive antifreeze, and in 1950 
this use consumed about 75 per cent 
of the total ethylene glycol produced. 

C. By using past figures for motor vehi- 
cle production and motor vehicle reg- 
istration as a basis, it is estimated that 



the total motor vehicle registration 
will be about 78,000,000 by 1965 and 
in excess of 100,000,000 by 1980. 

D. Over the past two decades, ethylene 
glycol has increased in favor over 
methanol as an automotive antifreeze. 
From 1947 to the present time there 
has been a rather sharp increase in 
the percentage of the antifreeze mar- 
ket controlled by ethylene glycol. At 
present, ethylene glycol is believed to 
account for about 55 per cent of the 
antifreeze market. 

E. Data relating to the per-vehicle con- 
sumption of antifreeze over the past 
20 years indicate considerable fluctua- 
tion. There are indications, however, 
that the consumption is stabilizing at 
a value of about 1.9 gallons per year. 

F. Assuming that piston-type engines 
are the only type engine used in mo- 
tor vehicles, requirements for ethylene 
glycol antifreeze should increase sub- 
stantially by 1965, to a figure of 
770,000,000—910,000,000 pounds an- 
nually. The above range depends on 
whether ethylene glycol continues to 
account for 55 per cent of the anti- 
freeze market or whether the portion 
of the market enjoyed by glycol should 
increase. . . . 



The researchers decided to recommend 
further studies to "firm up the figures 
presented in this report"; what areas 
should they suggest, and with what pri- 
orities? How much reliance should man- 
agement have in the present report as a 
picture of short-run trends? Long-run 
trends? Could it be improved without a 
large amount of additional effort? 



TABLE 2-R. PROJECTED MOTOR 
VEHICLE POPULATION 

(Thousands of Units) 





Passenger 


Trucks & 




Year 


Cars* 


Buses 
9,800 


Total* 


1955 


49,800 


59,600 


1956 


51,250 


10,140 


61,390 


1957 


52,710 


10,500 


63,210 


1958 


54,190 


10,880 


65,070 


1959 


55,670 


11,260 


66,930 


1960 


57,170 


11,650 


68,820 


1961 


58,690 


12,050 


70,740 


1962 


60,210 


12,460 


72,670 


1963 


61,750 


12,890 


74,740 


1964 


63,300 


13,320 


76,620 


1965 


64,860 


13,760 


78,620 


1966 


66,440 


14,210 


80,650 


1967 


68,020 


14,670 


82,690 


1968 


69,620 


15,150 


84,770 


1969 


71,230 


15,630 


86,860 


1970 


72,860 


16,120 


88,980 



♦Includes publicly owned vehicles. 

TABLE 2-S. CONSUMPTION OF ANTIFREEZE 1 
Methanol Ethylene Glycol 







Volume % 




Volume % 


MM Gals.* 


Year 


MM Gals 


of Total 


MM Gals. 


of Total 


Total 


1934 


5.9 


12.3 


4.0 


8.3 


48 


1935 


7.8 


15.0 


8.2 


15.8 


52 


1936 


10.1 


23.5 


9.0 


20.9 


43 


1937 


13.5 


27.6 


9.0 


18.4 


49 


1938 


11.1 


29.6 


9.0 


23.7 


38 


1939 


14.2 


32.3 


9.0 


20.5 


44 


1940 


17.4 


34.1 


11.0 


21.6 


51 


1941 


21.3 


28.8 


16.0 


21.6 


74 


1942 


2.8 


6.8 


13.0 


31.7 


41 


1943 


2.5 


3.7 


12.0 


17.9 


67 


1944 


8.0 


14.0 


13.0 


22.8 


57 


1945 


4.7 


9.6 


13.0 


26.5 


49 


1946 


20.0 


25.3 


19.6 


24.8 


79 


1947 


21.0 


26.9 


18.0 


23.1 


78 


1948 


40.0 


49.4 


27.0 


33.3 


81 


1949 


39.0 


48.8 


33.8 


42.2 


80 


1950 


41.0 


47.7 


40.0 


46.5 


86 


1951 


45.0 


46.9 


46.0 


47.9 


96 


1952 


47.0 


43.1 


58.0 


53.2 


109 



♦Includes, in addition to methanol and ethylene glycol, glycerine, ethanol, 
isopropanol, and propylene glycol. 

(1) Source: Chemical Economics Handbook by Stanford Research Institute. 
NOTES: (a) Methanol data for 1931 from Petroleum Processing, January, 1949; 

data for 1934-38 from Chemical and Engineering News, October 11, 

1948. Figures for 1949-52 estimated from reports in technical 

literature, 
(b) Ethylene glycol data for 1934-42 and 1945-50 are from Petroleum 

Processing October, 1951; data for 1943, 1944, 1951, and 1952 are 

estimates based on ethylene glycol production. 



105 









ETHYLENE 


GLYCOL — 








BU 




1 


% OF TOTAL ANTIFREEZE MARKET 




II 1 1 


1 1 1 














60 
















^. 


40 






































PERCENT | 


20 

10 


























8 




1 1 1 1 


1 1 1 1 


1 1 1 1 


1 1 1 



1935 



1940 



1945 
Figure 2-T 



1950 



GALLONS OF ANTIFREEZE 
PER REGISTERED VEHICLE 



4.0 




1930 



I960 



TABLE 2-V. PROJECTED ETHYLENE GLYCOL REQUIREMENTS 
FOR ANTIFREEZE-CASE A 



Year 



Registered 
Vehicles (MM) 



Total Antifreeze 
Required MM Gal 



Glycol Antifreezi 



MM Gal. 



MM Lbs. 



1955 


59.60 


114.5 


63.0 


584 


1956 


61.39 


117.9 


64.8 


600 


1957 


63.21 


121.4 


66.8 


618 


1958 


65.07 


125.0 


68.8 


637 


1959 


66.93 


128.5 


70.7 


655 


1960 


68.82 


132.2 


72.7 


674 


1961 


70.74 


135.8 


74.7 


692 


1962 


72.67 


139.6 


76.8 


712 


1963 


74.74 


143.4 


78.9 


731 


1964 


76.62 


147.1 


80.9 


750 


1965 


78.62 


151.0 


83.1 


770 


1966 


80.65 


155.0 


85.3 


790 


1967 


82.69 


158.8 


87.4 


810 


1968 


84.77 


162.8 


89.6 


830 


1969 


86.86 


166.8 


91.8 


850 


1970 


88.98 


170.9 


94.0 


871 



Assumptions: 

i. Antifreeze consumption— 1.92 gallons per vehicle, 
ii. Ethylene glycol constitutes 55% of market. 



106 



UNDERSTANDING THE CONSUMER 



107 



TABLE 2-W. 



PROJECTED ETHYLENE GLYCOL REQUIREMENTS 
FOR ANTIFREEZE-CASE B 



Registered 
Vehicles 
Year (MM) 



Total 
Antifreeze 
Required 
(MM Gals.) 



Glycol Antifreeze 



% of 
Market 



MM Gals 



MM Lbs. 



1955 


59.60 


114.5 


55.0 


63.0 


584 


1956 


61.39 


117.9 


56.0 


66.0 


611 


1957 


63.21 


121.4 


57.0 


69.3 


642 


1958 


65.07 


125.0 


58.0 


72.5 


671 


1959 


66.93 


128.5 


59.0 


75.8 


702 


1960 


68.82 


132.2 


60.0 


79.3 


734 


1961 


70.74 


135.8 


61.0 


82.8 


766 


1962 


72.67 


139.6 


62.0 


86.5 


801 


1963 


74.74 


143.4 


63.0 


90.3 


836 


1964 


76.62 


147.1 


64.0 


94.2 


872 


1965 


78.62 


151.0 


65.0 


98.2 


910 


1966 


80.65 


155.0 


65.0 


100.7 


933 


1967 


82.69 


158.8 


65.0 


103.3 


956 


1968 


84.77 


162.8 


65.0 


105.8 


980 


1969 


86.86 


166.8 


65.0 


108.4 


1,003 


1970 


88.98 


170.9 


65.0 


111.0 


1,028 



Assumptions: 

i. Antifreeze consumption— 1.92 gallons per vehicle, 
ii. Ethylene glycol portion of market as shown in Column 4 above. 



PROJECTED REQUIREMENTS FOR 
ETHYLENE GLYCOL ANTIFREEZE 



2.0 
1.5 


1 1 1 1 


1 1 1 1 


1 1 1 1 






CASE B 


MILLION LBS | 


1.0 










0.8 




___- — 1 




■ CASE A 


0.6 
















0.4 








1 1 1 1 


1 1 1 1 


1 1 1 1 



1955 



I960 



1965 



1970 



Figure 2-X 



108 



UNDERSTANDING THE CONSUMER 



CASE 2-8 

Bramton Company 



In late summer, 1953, Mr. W. T. 
Holderness, director of marketing re- 
search for the Bramton Company, was 
in the process of writing a report cover- 
ing a survey of the geophysical paper 
market which had been carried out under 
his direction. 18 He had been asked to rec- 
ommend whether or not Bramton should 
enter the geophysical paper market and, 
if so, whether the company should sell 
direct or distribute its paper through 
dealers. 

The Bramton Company manufactured 
a wide line of photographic films, papers, 
and supplies. Founded in 1927, it had 
grown rapidly, particularly during the pe- 
riod following World War II. This ex- 
pansion had been the result of both in- 
ternal growth, through reinvestment of 
earnings, and the purchase of smaller 
firms which made related products. In 
1948, Bramton had acquired the Davis 
Company, a small producer of photo- 
graphic papers. Prior to World War II, 
the Davis Company had been one of the 
principal suppliers of geophysical photo- 



18 Geophysical photographic paper is used in 
making geophysical recordings. The first step 
in making a recording is to place an explosive 
charge in the ground. The detonation of this 
charge sends a seismic shock into the earth, 
and the echoes created by the shock are 
picked up by geophones which translate the 
physical impulses to electric impulses. The 
geophones are disposed about the charge in 
numbers ranging from 6 to 24. The impulse 
from each geophone leads into an amplifier 
which builds up the intensity of the charge. 
From that point the impulse from each geo- 
phone goes to a mirror galvanometer. This 
galvanometer is activated by the intensity of 
the impulse causing it to fluctuate. As it 
fluctuates, it casts a reflected light on to sen- 
sitized paper which describes the trace lines. 



graphic paper. During the war, however, 
Davis had concentrated on other mar- 
kets and had discontinued its geophysical 
paper line. 

The Bramton Company maintained 
separate sales forces for its consumer and 
industrial product lines, and had branch 
sales offices in principal United States 
cities. 

In early 1953, the Bramton manage- 
ment had decided to examine the market 
for geophysical paper with a view toward 
deciding whether or not to enter the mar- 
ket. Within the preceding year, Bramton 
scientists had developed an emulsion 
which they believed to be very close to 
the improved emulsions then used on 
competitors' geophysical papers. This de- 
velopment, together with the known high 
rate of oil well explorations and the con- 
tinuing pressure from the branch mana- 
gers of the company's Dallas and Los 
Angeles sales offices for Bramton to add a 
geophysical paper to its line, had 
prompted the study. 

Bramton's marketing research depart- 
ment had been established in late 1952. 
The geophysical paper project was one 
of the first studies attempted by the 
Bramton Company internally. Previously, 
the company had regularly used outside 
consulting firms. Mr. Holderness and Mr. 
Johnson, Bramton's vice president in 
charge of sales and advertising, had de- 
cided to use Bramton's internal organi- 
zation for this study for a number of 
reasons. They believed that it would be a 
relatively easy task to get the facts on the 
geophysical photographic paper market, 
because they knew from the Davis Com- 
pany's prewar experience that this market 
was highly concentrated in the Dallas 



UNDERSTANDING THE CONSUMER 



109 



and Los Angeles area. In addition, the 
subject was highly technical. Both Mr. 
Johnson and Mr. Holderness agreed that 
some Bramton people had varying degrees 
of knowledge of the subject, that the 
study would be difficult and time-consum- 
ing, and therefore it would be expen- 
sive to educate a consultant's personnel. 

Preparatory to making the survey, Mr. 
Holderness checked on the distribution of 
geophysical exploration crews (see Table 
2-Y) and also determined which compa- 
nies were the leaders in this activity. He 
selected a sample of 11 companies with 
230 crews, or about 33 per cent of the 
total number of crews operating in the 
United States and Canada. Mr. Holder- 
ness found that a geophysical party used 
about a roll of recording paper a day. He 
believed that bad weather, which ham- 
pered explorations, would put the aver- 
age usage on an annual basis somewhat 
below 365 rolls for most crews, but that 
the heavier paper usage, required for geo- 
physical explorations over water, would 
raise the over-all average, so that "one 
roll per crew per day" could be stated as 
a rule of thumb. 

The results of the 11 interviews with 
geophysical exploration firms are sum- 
marized in Table 2-Z. In addition to the 
information set down on their interview 
sheets, the two researchers, on the basis 
of these interviews, gathered a number of 
impressions which they believed would be 
important factors affecting Mr. Holder- 
ness' recommendation: 

1. Pangar's Dominance of the Market 
— The general preference for Pangar was 
fairly recent in origin. A short time pre- 
viously, Global had been used very widely, 
but, in the view of the consumers, Global 
quality had slipped while Pangar's had 
improved and become the preferred geo- 
physical paper. 

2. Desirable Paper Characteristics — 
The importance of consistently good qual- 



ity could not be overemphasized. In the 
last analysis, all of the time and money 
spent in putting a geophysical party into 
the field resulted in only the photographic 
record. According to the consensus of the 
users, these were the specific character- 
istics which a good paper should have: 

Speed. A paper with an emulsion which 
would give a good contrast at the rated 
voltage of the bulb (about 4.2) when the 
paper traveled at an average rate of about 
12 feet per second, with a range of be- 
tween 7 inches to 17 inches per second. 

High wet strength. Under operating 
conditions it was often necessary to store 
the paper in water for several days before 
it was developed, and the wet strength 
became much more important than in 
conventional photographic uses. 

Lack of stain and mottle. The traces 
were often quite faint, and stains or mot- 
tles could obscure the light trace mark. 

No splices. Splices could not be toler- 
ated. In addition to the horizontal trace 
marks there was also a vertical time line 
which was used to orient the oscillations 
with respect to time elapsed. A splice 
would disrupt the proper timing sequence 
and cause serious difficulties in trace 
analysis. 

Wide processing latitude. The papers 
were processed in the field by men who 
knew little about photographic proce- 
dures. The crews operated in extreme 
climatic conditions and in temperatures 
ranging from below freezing to well over 
100 degrees. The developer temperature 
often could run to over 100 degrees. The 
developer and fixer were sometimes mixed 
with contaminated water or even salt 
water. In dry areas there might not be 
enough water to mix the solution in the 
proper proportions, and often there was 
no water available for washing. In view 
of these conditions, a rugged, foolproof 
paper was required. 

Base stock consistency. In view of the 



110 



UNDERSTANDING THE CONSUMER 



poorly qualified photographic personnel, 
the consumer would usually set up stand- 
ard galvanometer lamp settings for all 
crews, and variations in paper quality 
would cause weak or unintelligible traces. 
This seemed to be one of the most com- 
mon causes of dissatisfaction with sup- 
pliers and the changing of brands of 
paper. 

3. Buying Habits and Inventory Policy 
— A large proportion of the geophysical 
paper was purchased in Houston and Dal- 
las — probably better than 50 per cent of 
the United States consumption. Further- 
more, many companies had all of their 
paper sent to Houston where it was re- 
shipped to field crews even in areas as 
remote as the Northwest Territories in 
Canada. 

For many years, the consumers of geo- 
physical papers purchased direct from the 
manufacturers. More recently, however, 
they had gone to dealers for their needs. 
There were several reasons for this change. 
First, the paper manufacturers had in- 
sisted that the business be placed through 
dealers, probably as a means of strength- 
ening their dealer organizations. Second, 
the consumers experienced difficulty in 
getting prompt shipment from the manu- 
facturers, and delays of six weeks were 
not uncommon. In addition, purchasing 
was simplified because the dealers in geo- 
physical papers were also sources of other 
supplies which the consumer required. 
The place of the dealer in the distribution 
pattern seemed to be quite well estab- 
lished, particularly in the case of a small 
user or where purchasing was decentral- 
ized. These consumers had become accus- 
tomed to prompt shipment and keeping 
low inventory level. 

In addition to their dealers, both Pan- 
gar and Global had factory representatives 
contacting the users. They offered a great 
deal of technical assistance which a dealer 
was generally not equipped to give. 



4. Price and Competitive Conditions — 
Price cutting was prevalent. Consumers 
were being deluged with offers from deal- 
ers. One exploration firm had told Mr. 
Holdemess that five different Global deal- 
ers were soliciting its account and that one 
dealer was selling at such a low price that 
the cash discount was his only margin. 

Even by direct selling it was difficult to 
see how Bramton could offer a price ad- 
vantage to consumers. Pangar had al- 
ready classified some companies as dealers 
in order to offer price concessions. 

Most large consumers were getting 
their paper at about 30 per cent off list 
prices. Typical prices were approximately 
$10.45 list and $7.31 net for a roll 9" x 
200'. There were some cases where the 
prices paid by the consumer, particularly 
the smaller consumer, were higher. 

During one of his visits to an explora- 
tion company, Mr. Holderness was able 
to prevail upon the firm to conduct a 
field test of the paper Bramton was con- 
sidering marketing as a geophysical paper. 
The paper was tested and compared with 
the Pangar paper. 

In the case of the Pangar paper, the 
testers found that the proper exposure 
was achieved by running the galvanometer 
lamp at 4.2 volts. To achieve the same 
contrast with Bramton paper, the gal- 
vanometer lamp had to be run at 4.4 
volts. 

This was a disadvantage inasmuch as 
the increased voltage would diminish the 
life of the lamp and, consequently, would 
create the possibility of the lamp burning 
out during the run. However, the testers 
did find that, at the lower voltage of 3.8, 
Bramton paper showed a better contrast 
than Pangar; but as the voltage increased 
from 3.8, Bramton contrast became less 
pronounced than Pangar, and the Pangar 
seemed to pick up contrast a great deal 
faster. 

Unfortunately, at 3.8 volts in either case, 



UNDERSTANDING THE CONSUMER HI 

the contrast was not adequate for the ex- * * * 

ploration firm's purposes. TT7 , , . ,. , ,. , T 

r rr-, » lu £i. What recommendations should Mr. 

The testers other comments after run- T , , , , , * . .t.t, 

., . . £ n r> . » Holderness make? Assuming that Bram- 

ning the test were as follows: Bramton s . . „ ° r . . 

i . i i . i .. ton cannot match the Pangar finish or 

paper appeared to have somewhat better , , , . _ . & .. . . 

, rr ., ., r> rp, r • t contrast at 4.2 volts, even with further 

wet strength than Panear. The finish on . . ' .... 

r> t if .i r» . improvement by the Bramton scientists, 

Pangar paper was better than Bramton, . r . / , _ , ' 

, .v • . . i • v what chances does the Bramton product 

because there was a grain texture which _ , , m . f . _ 

j c r . • i • i have on the market? Assuming that the 

made for simphcty in making pencil no- _ _ b . , , 

tations on the seismic record. This was Bramton management does decide to go 

an important consideration inasmuch as a ^ ead wlth * ls P r ° duct ' , ho ^ *° uld ll 

• j i . _ . .i , plan to appeal to the market? What are 

engineers and analysts were constantly r rr 

making notations on the paper. The test- ^ e facto / s that determine demand for 

ers' opinion was that Bramton paper was «*« P^duct-price, service, or quality? 

not quite so good as Pangar but some- And how do they compare in importance? 

what better than Global. Who makes the h ^ m Z decisi0n? 



TABLE 2-Y. BRAMTON COMPANY 

Geophysical Activity 
March 31, 1953 

State Seismic Crews 

Arkansas 4 

California 14 

Colorado 16 

Illinois 1 

Kansas 11 

N. Louisiana 12 

S. Louisiana 106 

Total Louisiana 118 

Mississippi 26 

Montana 21 

Nebraska 1 

New Mexico 36 

North Dakota 12 

Oklahoma 35 

South Dakota 3 

East Texas 25 

North Texas 12 

Southeast Texas 43 

Southwest Texas 39 

West Central Texas 5 

West Texas 60 

Total Texas 184 

Wyoming _47 

Total United States 529 

Canada 179 

Total 708 



112 UNDERSTANDING THE CONSUMER 



TABLE 2-Z. BRAMTON COMPANY 

(Eleven companies interviewed; 232 crews operated, including 36 
Canadian crews) 

Number of 
Type of paper used Companies 

Pahgar 100% 9 

Pangar 95%, Triumph 5% 1 

Pangar 80%, Globe 20% 1 



1 
2 
2 

1 
2 
3 

1 
1 

1 
1 

2 

1 
2 

Purchasing Practices 

Purchase all requirements centrally from 

a dealer 5 

Purchase locally from dealers 4 

Purchase direct from Pangar (i.e., classified 

by Pangar as a dealer) 2 

Inventory Policy 

Minimum inventory— depend on dealer for 

fast service 7 

Maintain inventory of 6 to 10 weeks 4 

Equipment 

Manufacture all their own equipment 7 

S. I. E. camera 4 



Size 


of Rolls* 


6" 


x 100' 




6" 


x 150' 




6" 


x 200' 




8" 


x 100' 




8" 


x 150' 




8" 


x 200' 




10' 


' x 200 




15 


cm. x 


100' 


15 


cm. x 


150' 


15 


cm. x 


200' 


20 


cm. x 


200' 


25 


cm. x 


200' 


30 


cm. x 


200' 



♦Some companies used several sizes. 



UNDERSTANDING THE CONSUMER 



113 



CASE 2-9 

Southern Pine Paper Company 



In March 1952, the top executives of 
the Southern Pine Paper Company re- 
ceived a $35,000 report from Moody, 
Farrell & Nagel, New York management 
consultants. The report analyzed the ad- 
visability of the company's adding fatty 
acids and rosin derived from tall oil 19 to 
its line of products. 

The Southern Pine Paper Company 
was a leading manufacturer of paper and 
paperboard. A relatively small but grow- 
ing part of its business was derived from 
its chemical division, which sold one 
grade of refined tall oil. The company 
had entered the tall oil refining field, 
which utilized a by-product hitherto 
burned as fuel during World War II. At 
that time, markets were assured and ac- 



19 Tall oil is an oil obtained from the pine 
wood used in making paper. Its chief com- 
ponents are fatty and rosin acids. The Ameri- 
can Society for Testing Materials defines it 
as follows: 

"The natural mixture of rosin acids re- 
lated to abietic acid and of fatty acids 
related to oleic acid, together with non- 
acidic bodies which is the product of the 
acidification of the skimmings from the 
black liquor of the alkaline paper bulk 
industry." 

In making wood pulp by the sulphate pro- 
cess all fats, fatty acids, rosins and rosin 
acids in the wood are converted into soaps. 
These soaps are washed out of the pulp and 
collect in the waste black liquor. When the 
liquor is evaporated, the soaps separate as a 
brown, frothy curd, known as skimmings. 
About half of the latter is soap, the re- 
mainder being black liquor substances and 
water. 

Tall oil plants convert the skimmings into 
tall oil by mixing it with sulfuric acid. The 
mixture separates and an oil is drawn off. 
This is tall oil. Made in this way, without 
further processing, it is designated as crude 
or whole tall oil. 



celerated depreciation available for the 
fairly inexpensive equipment required. 

After the war, Southern Pine's tall oil 
continued to find applications, but sales 
of this product did not come up to the 
company's capability to produce it, which 
was expanding as its production of paper 
and paperboard expanded. 

Tall oil is about 50 per cent rosin and 
50 per cent fatty acids. The fatty acids 
are similar to those derived from cotton 
seed and soy beans. The grade of tall 
oil produced by Southern Pine was refined 
to remove some odor and color, but the 
two components were not separated. The 
chemical division's sales department had 
repeatedly urged that the company fur- 
ther process its tall oil to separate the 
rosin from the fatty acids. The sales de- 
partment argued that such a move would 
prove profitable in two respects: 

1. It would expand the market — many 
more customers would be interested in 
the fatty acids or the rosin alone than 
in the two in combination. 

2. It would permit the charging of an 
additional profit to compensate for 
the additional processing. 

Excerpts from the consultants' report, 
which totaled 104 typewritten pages, are 
presented herewith: 

FOREWORD 

The basic purpose of this study has 
been to supply a basis for determining the 
advisability of Southern Pine Paper Com- 
pany entering the field of producing fatty 
acids and rosin from tall oil. 

The attack on this problem has con- 
sisted of measuring the market available 



114 



UNDERSTANDING THE CONSUMER 



to these products — establishing the prob- 
able amounts of these products that can 
be sold and the price at which consumers 
would be attracted to their use. 

As part of this study, Southern Pine 
Paper Company's position as a supplier of 
refined tall oil has been borne in mind 
and the probable future of this product 
has been examined. 

Trained staff members of Moody, Far- 
rell & Nagel called upon representative 
companies that manufacture a wide va- 
riety of end products which include oils, 
fatty acids, and rosin. During these inter- 
views, samples of fatty acids and rosins 
supplied by Southern Pine Paper Com- 
pany, submitted as representative of the 
end products the company is considering, 
were exhibited. 

Discussions were obtained with top- 
level executives, production people, tech- 
nical and chemical executives, and pur- 
chasing agents. A total of 318 product in- 
terviews in approximately 170 companies 
were obtained. The companies studied 
represent a substantial proportion of their 
industries, as high as 75 per cent or 89 
per cent in some cases. 

CONCLUSIONS 

Based upon the findings arising from 

the study, it is concluded that: 

1. Refined tall oil — chiefly because of its 
characteristic odor, color, and viscosity 
— should, in the broad sense, be looked 
upon as a marginal material. It finds 
its markets principally on a price basis 
and as a substitute for other, more ex- 
pensive oils and acids. For these rea- 
sons, refined tall oil is subject to se- 
rious fluctuations in sales — ranging 
from an excessive demand when other 
materials are scarce and high priced 
(as during the war) to an extremely 
limited demand when competitive ma- 



terials are in plentiful supply and at 
low prices (as in the present period). 

2. The ratio of refined tall oil used to 
the total consumption of oils and 
acids, against which tall oil competes, 
cannot be expected to increase ma- 
terially. The natural growth of con- 
sumption of tall oil arising from 
increased output of oil-containing prod- 
ucts (as population and industrial ac- 
tivity grow) can be expected largely 
to be offset by trends toward the use 
of other materials (synthetics, resins, 
crude tall oil, and tall oil fatty acids). 

3. Among industries that consume refined 
tall oil, and in which no trend away 
from oil is found, there exists some 
trend toward tall oils containing lower 
ratios of rosin. To serve the market 
well, oils of at least several different 
proportions of rosin are required. 

4. Southern Pine Paper Company, in the 
unique position of supplying only one 
rosin-content grade of refined tall oil, 
can expect increasing difficulty in mar- 
keting the full output of its product. 

5. Southern Pine Paper Company, in an- 
ticipation of the ultimate problem of 
marketing its refined tall oil, should 
take steps to improve its position of 
service to the market. These steps 
could consist of: 

a. Making available at least two dif- 
ferent and additional tall oils (either 
acid treated or distilled) — one con- 
taining 20 to 25 per cent rosin and 
a second containing 30 to 35 per 
cent rosin — while continuing the 
production and sale of the present 
45 per cent rosin product; 

b. Separating tall oil into its fatty acid 
and rosin fractions (each relatively 
free of the other); or 

c. Combining both of these forward 
steps. 

6. In light of the interest in and com- 



UNDERSTANDING THE CONSUMER 



115 



puted size of the market for tall oil 
fatty acids and rosin of a quality ex- 
hibited during the study, it is con- 
cluded that Southern Pine Paper Com- 
pany would be wise to consider the 
establishment of facilities for the pro- 
duction of fatty acids and rosin derived 
from tall oil and, at the same time, to 
maintain the manufacture and sale of 
the present refined tall oil product. 

7. There exists a market which can ab- 
sorb approximately at least 100,000,000 
pounds of fatty acids 20 and a minimum 
75,000,000 pounds of tall oil rosin 
per year. (In 1951, the consumption 
of rosin is reported to have totaled 
798,600,000 pounds; in the same year 
the consumption of soya and corn oils 
for inedible products amounted to 
1,698,800,000 pounds— these figures 
are supplied to indicate the conserva- 
tive nature of the estimates of the 
available market.) 

8. Under present conditions of price in 
the natural oils and acid and naval 
stores industries, tall oil fatty acids 
would be attractive at approximately 
6 1 /2 cents a pound and rosin derived 
from tall oil at about 7 to IVi cents a 
pound. 

9. Assuming that Southern Pine Paper 
Company chooses to continue to serve 
its good refined tall oil customers — 
and that this market would absorb 
12,000,000 pounds of refined tall oil 
annually — there remains, it is under- 
stood, the equivalent of approximately 
12,000,000 pounds of refined tall oil 
to be marketed. 

This quantity could, it is understood, 



20 The projected total consumption figure 
of 89,180,000 pounds arrived at in the 
computations following has been rounded 
up to 100,000,000 here on the basis of ad- 
ditional use in industries not covered spe- 
cifically by the survey. 



be broken down into about 6,000,000 
pounds of fatty acids and 6,000,000 
pounds of rosin. 

These quantities are 6 per cent and 7 
per cent, respectively, of the conserva- 
tively computed available market for 
these products. 

At say, 6Y2 cents a pound for the fatty 
acids and the rosin, the potential an- 
nual sales volumes of these products 
would be in the neighborhood of 
$780,000 per year. 

For the purpose of arriving at an es- 
timated total sales revenue, the income 
from the sales to existing customers 
(12,000,000 pounds at, say, 5 cents a 
pound or $600,000) needs to be added. 
The resulting approximately $1,380,000 
can then be considered against current 
sales revenues, understood to total in 
the neighborhood of $700,000. 

Appropriate explanations, tables, and 
computations follow: 

Basis for Determining the Price of 
Tall Oil Fatty Acids at which Re- 
spondent Companies would be Inter- 
ested 

During discussions of the price of tall 
oil fatty acids at which active interest 
would be generated, the companies stud- 
ied spoke — for the most part — in terms 
of prices relative to soya oil or soya fatty 
acids of the alkyd grade. It was brought 
out again and again that tall oil fatty 
acids should permit the user a saving 
over the use of soya oil or soya acids, if 
the material is to make headway in its 
competition with soya. This is to say that 
oil fatty acids plus the necessary esteri- 
fying agent should, to be successful, un- 
dersell soya oil; or — in competition with 
soya acids — should be offered at a price 
less than the market price for that ma- 
terial. 



*£ 


O 


o 


S> s 


o 


o 


R 0) 


o^ 


<o 


5 *P 


"^ 


00 






13 
*e q § to 



o 


o 


* 


co 


oo 


o 


CM 


■^ 


o 




OS 






c-^ 




in 


o 

CM 


CM 





m o 



CO 


c- 


CM 


rf 


CO 


o 



co 



c/) 



p . 



1 

o 
•a 



^3 S 



•a 2 

^ co 



e£ 



eP 



o 
o 
o 
~ in 

CM ** 

CD 



tJ« i-l OS 

as os t- 



o 


o 


OS 


^H 


o> 


o 


o 


■<*« 


o 


CO 


CO 


o 


t-^ 


<o 


00 


•<*< 


Oi 


© 














oT 


oo 


Oi 


"*l 


Oi 


CM 


■* 


•^ 


1—1 




CM 


IA 












CM 





O 
O 


o 


c— 


CO 
CM 


3 

o 


00^ 


^ 




O^ 




CO 
OJ 


cm 

lT5 


CM 

00 



CO 


O 


c- 


CO 


o 


o 


CM 


•^ 


o 


o 


oo 


o 

o 


o 


CO 


Oi 


1-4 


Tj< 














ci 


oo 


1-1 


m 


T-t 


Oi 


CD 


■** 


CM 


y-H 


^ 


00 


i-t 










CM 



.2 B >> 



<« C(J 

a i 

a' 
§•2 



-to 
■♦«» 



'6 



£ 3 



S 

o 

g 
II 



I" 



,_ CJ 



<«> 5?" 



116 







o 


f 


.2 «o 


^ 


« 


d 


SS "C3 




rH 


,-■ 


o-c 








^ ^c 








Ci) 




O 


o 


* "S3 




O 


o 


3S « 


o 


lO 


o 


O Q 


^ 


o 


c~ 






■^ 


CM 






•Q ~^ 








£o 








^3 




o 


o 


*£ 


d 


o 


o 
o 




^ 


CM 


o 






CM 


CM 






co 


CO 




^ 


r~ 


OJ 


tJ 








c» 








cy 








05 








S 




o 


O 


^ 




O 


o 


O 


o 


o 




^ 


o 


o 



in 


OJ 


o 


CM 


o 
o 
o 


o 
o 





&$ 


CD 


OS 






in 


in 


e 








^ 








o 




o 


o 


to 




o 


o 




o 


o 


o 




^ 


CM 


o 






CD 


m 





o 


O 


o 


o 


o 


o 




o 


O 


o 


o 


o 


o 


o 


o 


o 


in 


o 


m 


o 


^ 


en 

00 
CM 


CM 

m 

CM 


CD 


CT> 


CO 


CO 






10 Si 

o 



p-3 CO 

o c 
a o 



* ex 

S * 
M O 

aJ cfl 

a- -a 
- a» 

T3 -^ 



3 O 

cr cj 



117 



118 



UNDERSTANDING THE CONSUMER 



3. REASONS FOR REJECTING THIS MATERIAL AS GIVEN BY 

COMPANIES THAT HAVE TESTED TALL OIL FATTY ACIDS AND 

THEN DISCONTINUED USING IT 

Oils and Fatty 
Acids Used 



Thousands 
of Pounds 



Total 



No reasons given 



162,325 
9,000 



100.0 
5.5 



Balance 

' 'Product usable, but priced too high. . ." 

Objectionable product characteristics-Total 

. .rosin content undesirable. . ." 

. .poor drying; iodine value too low. . ." 

. .poor color retention. . ." 

. .poor film durability. . ." 

. .objectionable odor. . ." 

. .rancidifies; iodine value too high. . ." 



153,325 



100.0 



08,150 


70.5 


45,175* 


29.5 


32,100 


20.9 


13,175 


8.6 


6,875 


4.5 


1,875 


1.2 


1,600 


1.0 


200 


.1 



♦Details add to more than totals because some respondents gave more than 
one objection. 



4. REASONS FOR NOT HAVING TESTED TALL OIL FATTY ACIDS 



Total 

'■' Price too high in relation to its probable 
value to justify cost of tests. . ." 



Oils and Fattv 
Acids Used 

Thousands 

of Pounds 9c 

5,457* 



3,478 



100.0 



63.7 



Tests not warranted because of expected 
objectionable characteristics of finished 
product: 

". . .poor drying because of low iodine value. . 
". . .would reduce film elasticity. . ." 
". . .objectionable odor. . ." 
". . .rosin content. . ." 

''Limited familiarity with tall oil fatty acids. . ." 
"Feel that product is not applicable; use stearic, 

hydrogenated fish, hydrogenated tallow, etc." 
"Use too little fatty acids to bother with testing. . 



720 

720 

30 

9 


13.2 

13.2 

.6 


991 


18.2 


201 
6 


3.7 
.1 



♦Details add to more than totals because some respondents gave more 
than one objection. 



5. DEGREE OF INTEREST IN THE SAMPLE OF TALL OIL FATTY ACIDS EXHIBITED 

Oils and Fatty Acids Used 









Substantial 


Moderate 


Slight, 


or 




Totat 




Interest 


Interest 


No Interest 




Thous. 




Thous. 


Thous. 




Thous. 






of Lbs. 


% 


of Lbs. % 


of Lbs. 


% 


of Lbs. 


% 


Total 


182,026 


100.0 


75,844 41.7 


93,876 


51.6 


12,306 


6.8 


Alkyd resins 
Wax and polish 


165,721 


100.0 


68,400 41.3 


88,215 


53.3 


9,106 


5.5 


emulsions 


2,039 


100.0 


2,000 98.1 


39 


1 9 






Liquid, paste, 
















and jellied 
















soaps 
Liquid buffing 


9,380 


100.0 


1,440 15.4 


4,740 


50.5 


3,200 


34.1 


compounds 


926 


100.0 


44 4.8 


882 


95.2 






Oiliness agents 


3,960 


100.0 


3,960 100.0 









6. INTEREST IN CONDUCTING TEST OF TALL OIL FATTY 
ACIDS AS EXHIBITED 



Total 



Oils and Fatty 
Acids Used 

Thousands 

of Pounds % 

182,026 100.0 



"Definite interest in testing, provided price 
is attractive and product is to be available 
in the market. . ." 

"Would test material, but primarily to keep 
abreast of developments. . ." 



133,011 73.1 

49,015 26.9 



7. REPORTED QUANTITY OF TALL OIL FATTY ACIDS THAT COULD BE USED 

ANNUALLY 

(By Companies Studies) 



Total 

\lkyd resins 
Wax and polish 

emulsions 
Liquid, paste, 

and jellied 

soaps 
liquid buffing 

compounds 
)iliness agents 

"■"Replaceable' 



Total Oils 
and Acids 

Thousands 
of Pounds 

182,026 
165,721 

2,039 



9,380 



926 
3,960 



Total 
"Replace- 
able" Oils 
and Acids* 

Thousands 
of Pounds 

115,107 
101,387 

1,438 



7,780 



Tall Oil Fatty Acids 



542 
3,960 



%of 
Total Oils 
and Acids 

63.7 
61.5 

69.2 
81.6 



442 
3,960 

oils and acids include soya oil, soya acid, oleic acid, and tall oil fattv acid. 



55.1 
88.8 



Potential 
Usage 

Thousands 
of Pounds 

55.535 



46,675 
238 

4,220 



%of 
Total 
Replace- 
able 
Acids 

48.3 
46.0 



16.6 



54.2 

81.5 
100.0 



120 



UNDERSTANDING THE CONSUMER 



8. COMPUTED 1951 TOTAL INDUSTRY CONSUMPTION OF 
"REPLACEABLE" OILS AND FATTY ACIDS 



Oils and Fatty Acids Used 





Total Fats 


"Replaceable" 






and Oils 
Thousands 


Fats and Oils 






Thousands 


%of 




of Pounds 


of Pounds 


Total 


Total 


289,000 


184,100 


63.7 


Alkyd resins 


252,000 


155,000 


61.5 


Liquid, paste, and 








jellied soaps 


19,000 


15,500 


81.6 


Oiliness agents 


8,000 


7,100 


88.8 


Wax and polish 








emulsions 


6,500 


4,500 


69.2 


Liquid buffing 








compounds 


3,500 


2,000 


55.1 



ESTIMATED TOTAL QUANTITY OF TALL OIL FATTY ACIDS 
THAT COULD BE USED BY INDUSTRIES CLASSIFIED AS 
OFFERING IMMEDIATE SALES POTENTIALS 

(1951 production basis) 







Reported 






Total 


Possible 






"Replace- 


Ratio of 


Potential 




able" 


Replacement 


Consumption 




Oils and 


by Tall Oil 


of Tall Oil 




Acids Used 


Fatty Acids 


Fatty Acids 




Thousands 


Thousands 




of Pounds 


% 


of Pounds 


Total 


184,100 


48.3 


89,180 


Alkyd resins 


155,000 


46.0 


71,300 


Liquid, paste, and 








jellied soaps 


15,500 


54.2 


8,400 


Oiliness agents 


7,100 


100.0 


7,100 


Wax and polish 








emulsions 


4,500 


16.6 


750 


Liquid buffing 








compounds 


2,000 


81.5 


1,630 



UNDERSTANDING THE CONSUMER 



121 



In respect to soya oil, the formula ex- 
pressing the "attractive" price of fatty 
acids derived from tall oil is found to be 
100 C fa + 12 C re + L = 104 C so - S 
in which: 

Cfa ~ cents per pound "attractive" 
price of the tall oil fatty acid, 

C re = cents per pound cost of the 
esterifying agent, 

L = labor cost of esterifying 100 
pounds of fatty acids and expressed in 
cents, 

C so = cents per pound price of soya 
oil, and 

S = a "savings factor," expressed in 
cents per pound. 

The "savings factor" — that is, the 
amount less than the soya price which 
would, respondents report, interest them 



in using tall oil fatty acids — ranges be- 
tween Vi cent and 2Vi cents per pound, 
as shown below. 

Reducing the above to a common de- 
nominator by taking 14 cents per pound 
as the price of soya oil (an approximate 
average price over the first quarter of 
1952), the weighted average "savings 
factor" is found to be 2.23 cents or 15.9 
per cent under the price of soya oil. 

In discussing the labor cost of esterify- 
ing tall oil fatty acids, the majority of 
respondent companies report that little, 
if any, additional cost would be incurred. 
Four companies claim that additional la- 
bor costs would be involved (in ordering, 
storing, and handling two materials — 
fatty acids and a vegetable oil — instead of 
one). See the lower table below. 



PRICE PER POUND LESS THAN SOYA OIL THAT WOULD 
INTEREST RESPONDENTS IN USING TALL OIL FATTY ACIDS 

Oils and Fatty 
Acids Used 



Total 



Thousands 
of Pounds 

118,785 



100.0 



5 per cent under soya oil 
10 per cent under soya oil 
15 per cent under soya oil 
18 per cent under soya oil 
20 per cent under soya oil 
33 % per cent under soya oil 

% cent per pound under soya oil 
1 cent per pound under soya oil 
1% cents per pound under soya oil 
2% cents per pound under soya oil 



1,500 


1.3 


15,980 


13.5 


23,600 


19.9 


1,250 


1.0 


13,490 


11.4 


20,000 


16.8 


4,975 


4.2 


18,670 


15.7 


12,000 


10.1 


7,320 


6.2 



REPORTED ADDITIONAL LABOR COST INVOLVED IN 

ESTERIFYING FATTY ACIDS RATHER THAN USING NATURAL 

ESTERS (SOYA OIL) 

Oils and Fatty 
Acids Used 



Total 



"No additional labor costs involved. 
$.80 per 100 pounds of acid 
$.12 per 100 pounds of acid 
$1.20 per 100 pounds of acid 
$.90 per 100 pounds of acid 



Thousands 
of Pounds 

156,695 



100.0 



132,735 


84.7 


9,300 


5.9 


7,840 


5.0 


5,500 


3.5 


1,320 


.9 



122 



UNDERSTANDING THE CONSUMER 



From the above, the weighted average 
additional labor cost is found to be 1.03 
cents per 100 pounds of fatty acids es- 
terified. 

Applying these values to the foregoing 
formula produces the following expres- 
sion: 

100 Cfa + 12 Cre + 1-03 = 104 C S o (-841) 



or 



Cfa 



87.46 Cso - 12 Cre - 1.03 
100 



Based upon this formula, the following 
table of "attractive prices" of tall oil fatty 
acids has been derived, at various prices 
of refined soya oil and of an esterifying 
agent. 21 



weighted average "savings factor" of 14.5 
per cent. (See table at right.) 

The weighted average "savings factor" 
from the above, computed on the basis 
of 12 cents per pound for soya acids, is 
found to be 1.83 cents per pound, or 
15.25 per cent. 

The formula expressing the "attractive" 
price of tall oil fatty acids in terms of 
soya acids would, of course, be: 

Cfa = Csa _ S, 

Cfa = cents per pound "attractive" 
price of tall oil fatty acids, 

C sa = cents per pound price of soya 
acids, 

S = "savings factor" expressed in cents 
per pound. 



'ATTRACTIVE" 



PRICE OF TALL OIL FATTY ACIDS RELATIVE TO PRICES OF 
SOYA OIL AND AN ESTERIFYING AGENT 









(Cents per pound\ 










Price Per 
Pound of 
Soya Oil 






Price Per Pound of Esterifying Agent 






25 P 


30 P 


35 P 


40 P 


45p 


50 P 


55 P 


60 P 


8<? 


3.99 


3.38 


2.79 


2.19 


1.59 


.99 


.39 




1CX? 


5.74 


5.14 


4.54 


3.94 


3.34 


2.74 


2.14 


1.54 


12C 


7.48 


6.88 


6.28 


5.68 


5.08 


4.48 


3.88 


3.28 


14? 


9.23 


8.63 


8.03 


7.43 


6.83 


6.23 


5.63 


5.03 


16C 


10.98 


10.38 


9.78 


9.18 


8.58 


7.98 


7.38 


6.78 


18<? 


12.73 


12.13 


11.53 


10.93 


10.33 


9.73 


9.13 


8.53 


20<? 


14.48 


13.88 


13.28 


12.68 


12.08 


11.48 


10.88 


10.28 


22? 


16.23 


15.63 


15.03 


14.43 


13.63 


13.23 


12.63 


12.03 


24C 


17.98 


17.38 


16.78 


16.18 


15.58 


14.98 


14.38 


13.78 


2&£ 


19.73 


19.13 


18.53 


17.93 


17.33 


16.73 


16.13 


15.53 



In respect to tall oil fatty acids as a 
replacement for fatty acids derived from 
soya oil, the companies studied report a 

21 For purposes of this case, assume that 
the current (1952) price of esterifying 
agents is running between 25^ and 30^. 



As discussed above, the weighted aver- 
age "savings factor" is found to be 15.25 
per cent. Applying this value the formula 
becomes: 

Cfa = .8475 C sa - 



UNDERSTANDING THE CONSUMER 



123 



PRICE PER POUND LESS THAN SOYA FATTY ACIDS THAT 

WOULD INTEREST RESPONDENTS IN USING TALL OIL 

FATTY ACIDS 



Same price as soya acids 
5 per cent under soya acids 
10 per cent under soya acids 
17 per cent under soya acids 
20 per cent under soya acids 
30 per cent under soya acids 

V 2 cent per pound under soya acids 
1 cent per pound under soya acids 
1V2 cents per pound under soya acids 
2V2 cents per pound under soya acids 



Total 



Oils and Fatty 
Acids Used 

Thousands 

of Pounds % 



86,370 



100.0 



660 


.8 


1,500 


1.7 


13,725 


15.9 


5,000 


5.8 


5,650 


6.5 


16,150 


18.7 


4,975 


5.8 


18,670 


21.6 


12,720 


14.7 


7,320 


8.5 





"Attractive Price 


arket Price of 


of Tall Oil 


Soya Acids 


Fatty Acids 


in Cents 


in Cents 


per Pound 


per Pound 


9<? 


7.63? 


10 


8.47 


11 


9.32 


12 


10.17 


13 


11.02 


14 


11.86 


15 


12.71 


16 


13.56 


17 


14.40 


18 


15.26 


19 


16.10 


20 


16.95 


21 


17.80 


22 


18.65 


23 


19.49 


24 


20.34 


25 


21.19 



Based upon this formula, the "attrac- 
tive" price of tall oil fatty acids, at various 
prices of soya acids (alkyd grade), would 
be as shown in the left column. 



On the basis of such data, is South- 
ern Pine justified in investing in facilities 
to produce tall oil fatty acids and rosin, 
provided the net return on investment 
(after the operating costs of separating 
them) is satisfactory at a selling price of 
6V2 cents per pound for the end prod- 
ucts, and provided the market for rosin is 
satisfactory to the same degree as the tall 
oil fatty acid market? Did Southern Pine 
get its money's worth in research? 






124 



UNDERSTANDING THE CONSUMER 



CASE 2-1 

New Era Specialty Paper Company 



In 1959, the New Era Specialty Paper 
Company was considering the manufac- 
ture of some kind of moisture vapor bar- 
rier (to keep moisture from affecting con- 
crete and foundation materials in construc- 
tion projects). 

To determine the competitive standing 
of vapor barrier materials currently used 
by building contractors and the relative 
impact of architects' specifications and of 
price and product characteristics, 22 a ques- 
tionnaire was sent to a sample of heavy 
constructors and a sample of light con- 
structors. (The researchers felt that there 
might be some significant differences be- 
tween the two types of contractors — those 
engaged in building structures worth 



$50,000 or more, and those engaged in 
building smaller structures. 23 

The returns from three successive mail- 
ings were as shown below. 

Use of materials was reported as shown 
on the top of p. 125 (since the successive 
mailings showed increasing proportions of 
nonuse, the researchers felt the returns 
could not be projected to the total uni- 
verse of contractors). 

Other findings are shown in tables on 
the following pages. 



What kind of paper should the New 
Era Specialty Paper Company consider 



Heavy Contractors 

Mailing No. % of Sample 

First 81 15.3 

Second 43 8.1 

Third _50 9.5 

Total 174 32.9 



Light Contractors 
No. % of Sample 
106 10.4 



76 
_89 

271 



7.4 
8.7 

26.5 



22 In actual market price at the time of the 
case, FRF was much higher than 6 mil 
film, and 4 mil film was just a shade higher 
than FRP. Roofing paper was also higher 
than 6 mil film, and not particularly well 
designed for use as a moisture vapor bar- 
rier. 

28 Sweet's Catalogue listed 1,039 names in 
heavy construction, so defined, and 17,311 
names in light construction. 



manufacturing, assuming it could produce 
it at prices competitive with other manu- 
facturers? (It is particularly interested in 
the future of fiber reinforced film.) What 
appeals should it use in promoting the 
new product? What other marketing de- 
cisions can management make in the light 
of what it has learned about demand for 
this kind of material? 



UNDERSTANDING THE CONSUMER 125 



Number of Heavy 


Number of Light 


Contractors 


Using 


Contractors Using 


29 




17 


59 




82 


27 




57 


12 




27 


2 




3 
5 


45 




_80 


174 




271 



Fibre Reinforced Paper. 
4 mil Clear Polyethylene 
6 mil Clear Polyethylene 

Roll Roofing 

Fibre Reinforced Film . 
Plastic Coated Paper . . 
No Use 

Total 



TABLE 2-AA. INFLUENCE OF ARCHITECT ON MATERIAL USED 
% of Time the Type of Material 



To Be Used Is Specified by 
Architect 

0% 


Heavy Contractors 
% Checking 

3.8% 


Light Contractors 
% Checking 

18.5% 


25 

50 

75 

100 

Total 


18.5 

26.2 

32.3 

19.2 

100.0 


13.9 
19.6 
19.1 
28.9 
100.0 


Number of Answers 


130 


173 



TABLE 2-BB. MATERIALS USED BY CONTRACTORS 



V Contractors 


Light Contractors 


% Using 


% Using 


22.5% 


8.9% 


45.7 


43.0 


20.9 


29.8 


9.3 


11.0 


- 


1.6 


1.6 


2.6 


_ 


3.1 


100.0% 


100.0% 


174 


271 



Material 

Fibre Reinforced Paper 

4 mil Clear Polyethylene Film . 
6 mil Clear Polyethylene Film . 

Roll Roofing 

Plastic Coated Paper -Moistop . 

Fibre- Reinforced Film 

Other— Felt, Curing Compounds 
Total 

Number of Answers 



TABLE 2-CC. PRICE VS. PRODUCT BUYING MOTIVES 

Heavy Contractors Light Contractors 

Reason for Choosing the Product % Checking % Checking 

Lowest Purchase Price 43.0% 25.0% 

Product Characteristics 57.0 75.0 

Total 100.0% 100.0% 

Number of Answers 128 176 



126 



UNDERSTANDING THE CONSUMER 



TABLE 2-DD. PRODUCT CHARACTERISTICS LIKED MOST BY BUYING MOTIVES 



A. HEAVY CONTRACTORS 



Product Characteristic Liked Most 





Durability 


Easiest 


Most 




Mat 'I. 


During 




to 


Rot- 




Used 


Construction 


Place 


resistant 


Total 




No. 


% 


No. 


% 


No. 


% 


No. 




(Those Buying Because of Product Characteristics) 






FRP 


13 


100.0% 


- 


- 


- 


- 


13 


4 mil 


7 


20.0 


23 


65.7% 


5 


14.3% 


35 


6 mil 


13 


40.6 


10 


31.2 


9 


28.2 


32 


Roll Roof. 


1 


25.0 


2 


50.0 


1 


25.0 


4 


FRF 


2 


33.3 


2 


33.3 


2 


33.3 


6 


Total 


36 


40.0 


37 


41.1 


17 


18.9 


90 




(Those Buying Because of Lowest Purchase Price) 






FRP 


15 


79.0% 


4 


21.0% 


- 


- 


19 


4 mil 


5 


15.6 


20 


62.6 


7 


21.8% 


32 


6 mil 


- 


- 


1 


100.0 


- 


- 


1 


Roll Roof. 


- 


- 


7 


77.8 


2 


22.2 


9 


FRF 


- 


- 


- 


- 


- 


- 





Total 


20 


32.8 


32 


52.4 


9 


14.8 


61 




(Characteristics Liked Most Regardless 


of Buying Motive) 




FRP 


28 


87.5% 


4 


12.5% 


- 


- 


32 


4 mil 


12 


17.9 


43 


64.2 


12 


17.9% 


67 


6 mil 


13 


39.4 


11 


33.3 


9 


27.3 


33 


Roll Roof. 


1 


7.7 


9 


69.2 


3 


23.1 


13 


FRF 


2 


33.3 


2 


33.3 


2 


33.3 


6 


Total 


56 


37.1 


69 


45.7 


26 


17.2 


151* 



B. LIGHT CONTRACTORS 



Mat 'I. 
Used 



Product Characteristic Liked Most 



Durability 

During 

Construction 



Easiest 

to 
Place 



No. 



% 



No. 



Most 

Rot- 

resistant 



No. 



% 



Total 
No. 



(Those Buying Because of Product Characteristics) 



FRP 


7 


63.7% 


4 mil 


25 


30.2 


6 mil 


27 


44.2 


FRF 


2 


40.0 


Plastic Paper 


1 


33.3 


Roll Roof. 


2 


28.6 


Felt 


- 


- 


Curing Comp'd. 


- 


- 


Total 


64 


36.3 



4 


36.3% 


35 


42.2 


20 


32.8 


2 


40.0 


1 


33.3 


3 


42.8 


1 


33.3 


_a 


100.0 


69 


39.2 



23 


27.6% 


14 


23.0 


1 


20.0 


1 


33.4 


2 


28.6 


2 


66.7 



43 



24.5 



11 

83 
61 

5 
3 
7 
3 
3 
176 



UNDERSTANDING THE CONSUMER 127 



TABLE 2-DD (Continued) 
(Those Buying Because of Lowest Purchase Price) 



FRP 


3 


60.0% 


2 


40.0% 


- 


- 


5 


4 mil 


5 


14.3 


15 


42.7 


15 


43.0% 


35 


6 mil 


3 


30.0 


5 


50.0 


2 


20.0 


10 


FRF 


- 


- 


1 


100.0 


- 


- 


1 


Plastic Paper 


- 


- 


1 


100.0 


- 


- 


1 


Roll Roof. 


- 


- 


1 


100.0 






1 


Felt 


- 


- 


- 


- 


- 


- 


- 


Curing Comp'd. 


- 


- 


- 


- 


- 


- 


- 


Total 


11 


20.8 


25 


47.2 


17 


32.0 


53 




(Characteristics Liked Most Regardless of Buying Motive) 




FRP 


10 


62.7% 


6 


39.3% 


- 


- 


16 


4 mil 


30 


25.5 


50 


42.3 


38 


32.2% 


118 


6 mil 


30 


42.3 


25 


35.2 


16 


22.5 


71 


FRF 


2 


33.3 


3 


50.0 


1 


16.7 


6 


Plastic Paper 


1 


25.0 


2 


50.0 


1 


25.0 


4 


Roll Roof. 


2 


25.0 


4 


50.0 


2 


25.0 


8 


Felt 


- 


- 


1 


33.3 


2 


66.7 


3 


Curing Comp'd. 


- 


- 


1 


- 


- 


- 


3 


Total 


75 


32.8 


94 


41.0 


60 


26.2 


229* 



* Higher than number of answers shown in Table 2-CC because some 
respondents checked several product characteristics. 

BIBLIOGRAPHY TO SECTION TWO 

Centers, Richard, The Psychology of Social Classes. New York: Russell & Russell, 
Inc., 1961. 

Clark, Lincoln H., Consumer Behavior (a collection of readings). New York: 
New York University Press, 1954. 

Dichter, Ernest, The Strategy of Desire. New York: Doubleday & Company, Inc., 
1960. 

Dusenberry, James S., Income, Saving, and the Theory of Consumer Behavior. 
Cambridge, Mass.: Harvard University Press, 1949. 

Henry, Harry, Motivation Research (discussion of use of motivation research in 
marketing operations). New York: Frederick Ungar, 1958. 

Katona, George, The Powerful Consumer. New York: McGraw-Hill Book Com- 
pany, Inc., 1960. 

Martineau, Pierre, Motivation in Advertising: Motives That Make People Buy. 
New York: McGraw-Hill Book Company, Inc., 1957. 

Newman, Joseph W., Motivation Research and Marketing Management (includ- 
ing several detailed cases). Boston: Division of Research, Harvard Business 
School, 1957. 

Warner, W. Lloyd et al., Social Class in America. New York: Harper & Brothers, 
1949. 

See also bibliography to Section Four. 



SECTION THREE 



Product 
Potential 



In this section we shall concentrate on the question of product potential: how 
big or how good the market is for a given product or service. (The strategy of 
product introductions will be discussed in Section Six.) 

Note that the question includes both "how big" and "how good"; they are 
not necessarily the same thing. Thus, a new shade of lipstick may be so attractive 
that it wins 10 per cent of the market just through display, or it may be so radical 
an innovation that it takes a lot of promotion or retailer push to bring about a 
change in women's attitudes— yet, with that aid, it too wins 10 per cent of the 
market. Both markets are of the same size, but the first one develops more easily 
or inexpensively. In other words, market potential is not something static or fixed, 
but reflects qualitative influences like everything else in marketing. 

This question of product potential is the province of marketing research. The 
phrase "marketing research," rather than "market research" is used here because 
we are concerned with research as a tool for all the active aspects of marketing 
as it moves goods into people's hands and moves people to put their hands on 
goods— not just a still picture of the market (which doesn't "stay put," anyway). 
In the text part of this section we shall deal principally with marketing research 
and what it can do for management as an information tool; in a number of the 
cases at the end of the section, however, we shall go on to tackle some decisions 
that have to be made after the research findings are in. 

(Because of its complexity, much of the subject matter in the following 
pages is not treated in the full depth it deserves; since we do not have all the moot 
points footnoted, readers are encouraged to make more-than-usual use of the 
bibliography at the end of this chapter. ) 

129 



130 PRODUCT POTENTIAL 



MARKET TESTING 



The closer (in terms of time and realism) that one gets to the actual buying 
act in making measurements of product potential, the less slippage from the true 
amount of potential there is likely to be. This is the ideal, but practical marketers 
must often settle for less. 

Almost like the buying act would be a market test or store test, simulating as 
closely as possible the actual conditions and influences likely to affect buyers 
when the product is offered for sale. This is likely to be a very effective method, too, 
if it can be arranged properly. 

Since the product is going to be offered with certain characteristics, in a cer- 
tain kind of outlet, at a certain price, and with a certain kind of promotion, the 
variable being tested either can be the whole package of these factors (how well 
will the product, as totally envisioned, sell), or it can be any one of them (how 
well will the product sell at this or that price, with this or that advertising theme or 
degree of promotional effort, and so on). Accordingly, the procedures to be de- 
scribed apply not just to measuring market potential in the event of introducing 
a new product, but to predicting the outcome of any marketing move. 

Despite all the advantages of market testing, there are offsetting difficulties 
that must be taken into account: 

(1) It is difficult to isolate the variable being tested. Suppose it is an ad- 
vertising theme for a new product. It is tried in a TV commercial, and sales de- 
velop faster than with some other theme, or than with no advertising at all; pre- 
sumably the net gain is the result of that advertising. 

But how sure can one be that it was the effect of the advertising on the 
consumers, and not something else like attention given the product by retailers 
under the circumstances? You may not care what causes the increase in sales, but 
for long-run prediction you may be misled if the retailers' interest is only aroused 
by initial curiosity that will not last. Or it might be that, in the geographic 
area in question, a large plant has just put on an additional shift, and so more 
money is circulating, in which case the rise of sales might mean nothing about 
the effectiveness of the advertising. 

(2) One way to avoid some of the possible distortion caused by the effect 
of variables other than that being tested is to have a control: an area or panel of 
stores as much like the test area or store panel as possible. Thus, if factors like 
season of the year, weather conditions, or general economic conditions vary, hope- 
fully they will vary the same in both locations, and so the effect of the marketing 
move will be the difference between what happens to sales in the test area and 
what happens to sales in the control area. 

But even with this kind of test one may have difficulties. It is almost impossi- 
ble to get control and test conditions to match exactly. Further, both sets of con- 
ditions can become contaminated as the test progresses. People come into the test 



PRODUCT POTENTIAL 131 

area who aren't in range of the TV, or people in the control area happen to pick 
up the TV broadcasts or hear about the product from friends who live in the 
control area. Or, even worse, it sometimes happens that a competitor, either 
accidentally or deliberately, takes this particular moment to do some extra-strong 
promotion of his own. 

(3) Since this is a test of actual market reaction (rather than of attitudes 
that may lead to buying regardless of current circumstances), it is imperative that 
the circumstances be realistic. Yet it is difficult to simulate the effects of national 
promotion or effects over time, though for many products and many marketing 
actions both of these sets of effects are important. 

(4) Testing of this kind requires cooperation on the part of stores, or media, 
or whatever marketing institutions are involved. It takes considerable effort to 
make the arrangements and then to audit the results. Advantage can be taken of 
existing test arrangements such as maintained by A. C. Nielsen and Company 
and Market Research Corporation of America (see Section Five and Case 5-4); 
this of course involves a considerable monetary outlay, but may be worth every 
penny of the expense, depending on the sales volume of the company and the 
cruciality of the test result. Perhaps, even more serious, all this means delay, and 
management may not want to wait so long for a decision— either because of pres- 
sures for actions inside the company or for fear of competitors' stealing a march. 

It should be pointed out that testing does not need to be done only in ad- 
vance of (and for the purpose of) specific decisions. Rather than holding up op- 
erations while the best move that can be devised is being tested, why not, as a 
regular procedure, go ahead naturally with what looks like the thing to do— but 
always, just to build up data that can be converted into better action ultimately, 
hold out one or two territories or spots where the action is not taken and measure 
the differential results? This avoids the onus of having everything depend on one 
project and also minimizes the wrench to the rest of the organization, which may 
hesitate to have any interference with an important marketing move. The recur- 
ring tests can be much smaller, as well as more casual, yet in total add up to more 
significance, at less cost. Such a program obviously applies only to companies with 
a continuing research effort; and it is not suitable for new problems of unusual 
character or magnitude. 

In any event, experience indicates that management's fears of the conse- 
quences of delay are often exaggerated, and if the move being tested is a crucial 
one, a good, slow decision may cost a lot less than a poor, fast one. And this 
again is a question that only management judgment can decide. 

There are any number of ingenious variations on testing, such as: 

1. Use of split runs (where a magazine or newspaper arranges to carry an ad 
or a different ad in only certain copies) . 

2. Checkerboard designs (where the segments of a large number of sales terri- 
tories may be alternating test and control areas for one or several variables at the 



132 PRODUCT POTENTIAL 

same time; thus, in one test, the effects of different degrees of advertising, direct 
mail, and field selling could be measured. 

3. Multiple cross-comparisons (where the effect of small or minor variations 
can be detected by a massive number of tests, say, between a and b, b and c, a 
and c, analyzed in such a way that b-c is the control for a, a-c for b, and a-b for c; 
see Case 4-3. 

But in each case the point is to measure the result of marketing action under 
real conditions. And success depends on (a) the accuracy of measurement (not 
always easy to achieve when you must depend on other people's records or on 
audits before and after the event), and (b) the realness of conditions (which can 
be "loused up" by events not under your control). Knowing all this, management 
must decide when and how elaborately it wants to do market testing— which, 
literally, is the ultimate in marketing research. 

EXPERIMENTAL DESIGN 

Market testing owes its effectiveness to the fact that it is a very clear-cut 
variety of experimental design. So let us take a broader look at experimental de- 
sign, which consists of testing the effect of a variable under controlled conditions. 

First of all, it should be made clear that "controlled" does not necessarily 
mean that the conditions of the experiment are under the researcher's actual 
control in the sense that he can set them up the way he wants them to be, as 
a chemist does in his own laboratory. For the most part he has to deal with a slice 
of the world as it is. 

But "controlled" does mean that, by careful planning, the researcher can 
select the existing conditions that will serve his purposes and then arrange what 
he does so that he gets fruitful comparisons (or contrasts) from them. Here are 
the essential elements : 

(1) Proposition— the marketing action or idea that is to be tested. 

(2) Comparison— measurement against a control to test the proposed action. 

(3) Planning— purposeful design to make the desired comparison possible. 
Let us take these three elements in order. 



Proposition 

A proposition is a broader, more operational way of expressing a variable. 
Thus: "With this characteristic (the variable), the product will win a satisfactory 
share of the market"; or "At this price (the variable) a larger number of people 
will buy it than at some other price or prices"; and so on. It is not only helpful 
to have a proposition express single, simple variables of marketing action like 
product feature or price, but a proposition also accommodates itself to more 
comprehensive or complex ideas like: "Our customers think our products are 



PRODUCT POTENTIAL 133 

reliable (or unreliable)"; or, directly in line with our concern with product poten- 
tial, "Our strategy for this new product will be successful (or will result in x 
dollars of sales volume, or x per cent of return on investment)." 

Note that some researchers use the term "hypothesis" to describe what is to 
be tested, but from the viewpoint of marketing management, "proposition" has a 
happier connotation of action. "Hypothesis" can also be used to describe a re- 
search assumption. Thus, in a research project undertaken for an automobile 
manufacturer, Alfred Politz assumed that automobiles that are styled more to 
consumers' liking are more noticed and better remembered; and on the basis of 
this research assumption he measured styling differences by doing a mass survey 
of how many of the various makes of cars people thought were in their neighbor- 
hood, and compared this with actual registrations. 1 

In this book "hypothesis" will be reserved for such a use, that is, as an as- 
sumption for research procedure. In contrast, the "proposition" that Politz was 
researching could have been phrased: "The Brand X car is not as well-styled as 
competitors' cars." Confirmation of this would lead directly to marketing action 
on the part of management. He did find his client's car remembered far less than 
its actual incidence— and advised a sharp change in styling, with a highly bene- 
ficial effect. 

However, no matter whether a marketer uses the term "proposition" or not, 
his careful crystallization in words of what is to be tested may be the one most 
useful thing he can do before engaging in research. Not that the words are neces- 
sarily important, but the prethinking that they require is. 

Comparison 

As for comparison, this is the device that squeezes out predictive information. 
Usually, it turns out to be contrast. Thus, in the case of test marketing, it is con- 
cerned with the difference between the situation before a marketing move is made 
and after it is made; or, with greater refinement, the difference between (a) the 
difference between before and after in the test area and (b) the difference between 
before and after in the control area— for example, the difference between two rates 
of sales increase over the same period. Comparison focuses on the difference, 
measures it, and provides the basis for prediction or inference, which in turn be- 
comes the basis for action. 

But there are all sorts of comparisons possible. Here are four main types: 
TYPE I. Yes— No comparisons about a single product or about some one 
aspect of the marketing of a product. Thus: "Yes— I will buy," "I intend to 
buy," "I like the product." Or, "Yes— I am attracted by that promotion," "I think 
the price is right," "This is the kind of store where I expect to find it." And so 
on. These reactions may be in response to questions, or revealed by observation 

1 Alfred Politz, "Science and Truth in Marketing Research." Harvard Business Review, 
January-February 1957, p. 119. 



134 PRODUCT POTENTIAL 

(as in market testing) or by other indication, but the point is that, in each in- 
stance, they are compared against the corresponding "No." 

A variation is this-or-that comparison, where people's attitudes or dispositions 
toward several alternative products or aspects of a product are compared against 
one another. "I like this product better than that one" or "I find this advertising 
appeal more persuasive than that one, or that one." Here the various alternatives 
serve as controls for each other. 

Sometimes the comparison can be even more informative if it is made against 
an independent bench mark— either something to which the response is already 
known through past experience, or what buyers indicate they feel is ideal. For 
example, a pharmaceutical manufacturing company was pleased to find that doc- 
tors expected its product to be more "potent" than competing products, par- 
ticularly because the company scientists had purposely made the drug more 
potent, so a smaller dosage could be prescribed; they presumed that this was a 
desirable characteristic, since a lower dose should be safer to prescribe. But when 
the research was repeated, and doctors were asked to rate several competing drugs 
and the ideal drug for the disease in question, they rated the ideal as low in 
potency. The reason, as subsequent research revealed, was that potency had a 
connotation— illogical but psychological— of dangerous side-effects. 

TYPE 2. Comparisons among different kinds of people in their reactions 
to a product or aspect of marketing a product. These comparisons are made by 
looking at demographic differences like age, income class, education (which 
usually runs parallel to income class), occupation, social class, and geographical 
location, and at differences in personality characteristics like conservative, innova- 
tive, and so on. The purpose is not just to be able to project findings quanti- 
tatively in terms of proportions of the total population, but actually to produce 
further inferences about cause and effect. 

For example, if a luxury product elicits a much better response among high- 
income families than among low-income families, then it may be inferred that the 
price involved is the cause of the difference. Again, if a higher proportion of mid- 
dle-age women have a negative attitude toward a household product than do 
younger women, then something about the way the product performs may be alien- 
ating women the more they use it, or— as in a piece of research which deliberately 
set out to test this proposition in connection with use of zippers in home sewing 2 
the way most older women were trained to sew in their youth may make sewing 
seem an unpleasant duty, whereas today's younger mothers have been mostly 
brought up to regard it as fun. 

Or, to incorporate a this-or-that comparison, if factory workers, truck drivers, 
farmers, laundry women, and so forth, tend to prefer beer, and lawyers, architects, 



2 See Joseph W. Newman, Motivation Research and Marketing Management, Boston 
Division of Research, Harvard Business School, 1957, pp. 264 ff. 



PRODUCT POTENTIAL 135 

business executives, club women, and the like tend to prefer champagne, then it 
can be inferred that social factors affect the reactions of people to the beverages 
in question. 

Another very simple but useful kind of difference, in making comparisons 
among people, is between users and nonusers, between frequent users and occa- 
sional users, between present users and former users (that is, those who have 
discontinued using). For example, if those who frequently attend professional 
games tend to live in areas where telecasts of such games are available, whereas 
those who attend less frequently live in areas where the telecasts are not available, 
then it can be inferred that the telecasting of the games has a causal relationship 
to attendance. (Without an actual test, however, one can sometimes be fooled as 
to which is cause and which is effect. For example, do people who live in suburban 
areas like outdoor cooking and gardening because they live there or do they live 
there because they like such outside activities?) 

TYPE 3. Comparisons with similar products or similar situations , whereby 
inferences are drawn on the basis of "This new product or this new marketing 
move is like that one, and therefore will perform similarly." In this kind of ap- 
proach, it is the other product or other marketing move that serves as a control. 

This is one of the few uses where comparison is not essentially contrast; that 
is, where it is focusing on similarities rather than on differences. But this fact itself 
suggests a word of caution; namely, that anybody following this approach had 
better try hard to determine whether there aren't some overlooked differences that 
might be misleading if not taken into account. 

Thus, when Polaroid cameras were first brought out at $89.75, it would have 
been a mistake to expect them to sell the same as regular high-price cameras; the 
Polaroid, with its one-minute finished picture, appealed not to camera "bugs" and 
professionals, but to "dubs" and those interested in novelties. It is not the phys- 
ical characteristics of the product that make useful similarities, but the way people 
use it and think of it. Tomato ketchup may be more like mustard than like to- 
mato soup. Color TV today (you can live without it) is not like black-and-white 
TV used to be (you had to have it to be socially comfortable). 

Another diEculty is that what happens to a product is not just the result of 
what the product is, or of how people use it and think of it, but also of what 
marketers are doing to promote it, and of the circumstances which may or may 
not facilitate the effectiveness of such promotion. Why should the shape of the 
sales curve of the automatic clothes dryer, when first introduced, be like that of 
the automatic clothes washer in its introductory phase? The dryer ought to ac- 
celerate faster, since the market is more ready and the promotion "gets through" 
better. 

But, with care, this kind of comparison can be a useful device, simply be- 
cause it relies on something that has already happened. Less dependable, but 



136 PRODUCT POTENTIAL 

still useful, is a comparison of the conditions under which another product or 
marketing move succeeded (or failed to succeed) with the conditions under which 
the contemplated move is to be made. For example, an expensive household ap- 
pliance is more likely to make its way when introduced into a foreign country if 
conditions there are similar to what they were in this country when it was intro- 
duced here. Significant in this connection are disposable income, standard of liv- 
ing, family size, and electrification, as well as social trends and domestic customs. 
Or the similar conditions could be reactions like those discussed in Types 
1 and 2 above. When the particular degree or particular quality of people's reac- 
tions to the product are like the reactions to some other product, then the same 
or similar success can be expected of this product. Thus, when housewives show 
the same interest in the promotion for a new detergent as they did in the promo- 
tion of something else sold at about the same price and with the same frequency, 
then the new detergent has that much more likelihood of repeating the same 
market performance. But at this point, with this kind of comparison, we are pretty 
far away from cause and effect; we are looking at the shadows of causes. 

TYPE 4. Comparisons over time, whereby inferences can be drawn and pre- 
dictions made by observing different stages in the progress of cause and effect. 
Thus, when successive measurements of consumers' attitudes toward a product 
are compared, one can see the product taking hold in the market, and draw trend 
lines to predict the future. 

As shown in Case 5-6, there often are patterns that can be identified from 
past experience (to this extent we are using the No. 3 type of comparison, that is, 
similarity). It makes a great deal of difference whether a developing market con- 
tains a growing number of buyers who, having bought once, continue to buy, or 
is a fluid meld of ever more new buyers replacing old buyers who cease buying. 
In short, it makes a great deal of difference whether it is a true buyers' or only a 
triers' market, or something in between. 

In effect, with this type, we are looking at products (or the marketing of 
products) and at people as in the first two types of comparison, but with a new 
dimension added: time makes the comparison a moving picture, not just a 
snapshot. 

The simplest form of time comparison would be something like this: 

On April 1, 20 per cent of the market showed a buying intention for 
this product; on July 1, the figure was up to 30 per cent. Or it could be that, 
dividing our potential market into high-income and low-income, we see in- 
tention in the high-income segment holding steady at 26 per cent-27 per cent 
in total while in the low-income segment it has risen from 17 per cent to 31 
per cent in total. This double comparison (by people's income and by time) 
suggests that price resistance is being overcome. 

Then, adding a third comparative element (two different product designs), 
we might get: 



PRODUCT POTENTIAL 137 

PERCENTAGE HAVING BUYING INTENTION FOR: 
Higher-Price Design Lower-Price Design Both Designs 





April 1 


July 1 


April 1 


July 1 


April I 


July I 


High-Income 














Group 


16% 


13% 


10% 


14% 


26% 


27% 


Low-Income 














Group 


6 


16 


11 


15 


17 


31 


Total Market 


9 


15 


11 


15 


20 


30 



This clinches the price behavior, and further suggests that the larger fami- 
lies that need the larger size, being also the low-income families, are beginning 
to buy at an increasing rate. (This could be confirmed by making a com- 
parison by family size or maybe by type of dwelling). Then, if a No. 3 com- 
parison could be found, with a similarity to another product, the trend line 
could really be drawn with some confidence. 

Time comparisons can be refined in many ways— but not necessarily im- 
proved in every instance, because sometimes grosser measurements are all that 
are needed, and then more subtle measurements may represent a loss in usefulness 
because they are almost always less precise. 

One such refinement would be to use different measurements which them- 
selves contain a time or stage-of-development factor. Think of people as moving, 
from not being aware of a product to becoming aware, then interested, then 
feeling a preference for this product over some other product or other use of the 
money, then intending to buy, and finally buying— not in discrete stages like that, 
but in a smooth (though not necessarily constant) progression. It follows that 
comparing some of the indicators of such a progression, one can get a picture at 
one instant of how far the marketing program is in terms of its development over 
time, and therefore how it is likely to progress in the future. 

Thus, a picture of the market might be: 35 per cent are unaware of the 
product; 30 per cent are aware, but no more than aware; 20 per cent are interested; 
10 per cent have buying intentions; 3 per cent have bought, but not repeated; and 
2 per cent have bought and have become regular users. This comparison of how 
many people are at what stage suggests that sales are going to rise, and that 
promotion is making headway. Of course, the time dimension can also be com- 
pounded, and we can compare stages of development at several times. For example: 

January 1, 1961 January 1, 1962 Difference 



Unaware 


35% 


10% 


-25% 


Aware 


30 


20 


-10 


Interested 


20 


40 


+ 20 


Buying intention 


10 


20 


+ 10 


Triers 


3 


4 


+ 1 


Regular buyers 


2 


6 


+ 4 



Total 100% 100% 



138 PRODUCT POTENTIAL 

Or, piling dimension on top of dimension, we could make the same time 
comparisons but apply them to different kinds of people reacting to different 
attributes of the product; and— most significantly— inside the different kinds of 
people in terms of socio-economic characteristics, we could look at frequent and 
occasional users. 

Note that, at this point in our sequence of types of comparison, we have 
moved far from directly observing causes as they produce the end effect of the 
buying action; we are working indirectly, measuring intermediate effects that only 
imply the causes and (hopefully) the end effect. But also note here, as throughout 
the whole sequence, it is comparison of one factor with some other factor or fac- 
tors (that is, some control, in terms of product or people or experience or time) 
which brings out the inferential, predictive meaning. 

Planning 

We come now to the third essential element of experimental design— plan- 
ning. Let us look first at the planning of what kind of research to do. 

Some research is exploratory and consists of searching around in a market or 
in people's minds for incidents or ideas that will suggest lines along which fresh 
marketing action can be taken, or propositions on which more purposeful market- 
ing research can be done. Thus, for a pharmaceutical company interested in 
bringing out a new tranquilizer, researchers could simply sit down with doctors 
and try to draw them out as to how they felt about such drugs. In so doing, they 
might find a few doctors that feel uneasy about doing anything that seems to 
tamper with people's minds, or maybe a few that are afraid of the "slowing-down" 
effects of tranquilizers. 

Such exploratory (sometimes called "descriptive") research may be useful 
in just that form, but it can be enhanced if the company goes on to do testing of 
the planned or designed kind to test the proposition thus developed— to find the 
prevalence or strength of the feelings that have been discovered, how they can be 
countered or taken advantage of in promotion, and all the other ways they would 
affect the course of future sales. One company, on this basis, went on to test 
whether doctors could be sold on the idea of deliberately slowing down cardiac 
patients with tranquilizers, thus attempting the "home run" feat of marketing: 
turning an objection into an advantage. 

But even in exploratory research it seems a little wasteful to have no plan for 
what you are trying to find. Because you are casting your net wide and deep, you 
may not be able to afford to do it often enough (that is, get a large enough sample) 
so you can project quantitative figures; besides, the main objective is a good 
variety of interesting fish. Yet, even here, why not try to get more variety by 
purposefully designing comparisons? For example, why shouldn't the pharma- 
ceutical company described above plan to draw out at least what it thinks may be 
meaningful differences? Even if only 25 doctors are going to be interviewed, the 



PRODUCT POTENTIAL 139 

stage can be purposefully set for as many as six comparisons. Thus it would be 
possible, and might turn out to be highly useful, to have (1) both some old 
doctors and some young doctors, (2) both some general practitioners and some 
specialists, (3) both some county doctors and some city doctors; and to ask them 
questions (4) about strong and mild tranquilizers, (5) about tranquilizers and 
barbiturates (another drug that could be used for somewhat the same purpose), 
or (6) about tranquilizers and psychic energizers (another drug affecting mental 
attitude but with completely opposite purpose). 

The only danger here is that what started as a fishing trip may, just because 
you try to see what you can infer from the variety of fish, be taken as providing 
a census of marine life or as information justifying the building of a factory for 
canning seafood. Usually it is necessary to go further and test the inferences you 
have drawn on a larger, more rigorous sample. In fact, there is much to be said, 
not only for planning the design of any single experiment so as to make it most 
likely to be productive of meaning, but also for planning the design of a series of 
tests so that in total the company has the best probability of getting the informa- 
tion it needs at reasonable expense. Thus, one could have this planned sequence: 

1. Exploratory search— small-scale search for whatever fish can be found, 
but with some care to fishing in the waters where fish are known to be most 
prevalent (and maybe, also, a deliberate side-excursion to test the waters that 
others have taken for granted as being poor fishing); this provides the propo- 
sition for the main test. 

2. Pretest of main test— just a moderate number of observations, inter- 
views, or mailed questionnaires (it can be as low as a half-dozen observations 
or interviews or a hundred mailed questionnaires), with the purpose of find- 
ing out whether the instrument you are going to use has any "bugs" in it 
that need correction— as it always has. People always find some occasion for 
misunderstanding questions, or lapse of memory, or failure to maintain in- 
terest, or something. Indeed, it is almost categorical that a pretest is worth 
the time it takes. 

3. The main test— purposely designed, most often with comparison as 
the tool, to test the main propositions that management needs to know about 
in order to take intelligent action. (This kind of planning will of course be 
discussed in more detail later.) 

4. Follow-up of main test— to fill in the nooks and crannies that usually 
develop, particularly puzzling inconsistencies in the way some respondents or 
subjects of observation seemed to react (there often is a clue to something 
valuable in such an inconsistency), and, most important of all, checking on 
nonrespondents or those not covered by the test to make sure their "absence" 
does not distort the results. 

5. Action test or tests— for the purpose of (a) verifying or carrying further 
some of the subtler points of the main test findings; (b) testing action plans 



140 PRODUCT POTENTIAL 

based on the main findings (for example, advertising copy or salesmen's pres- 
entations); and (c) pushing the testing closer to the actual buying act or 
doing it on a larger, more definitive scale (for example, store testing or field 
testing if it has not already been done). And note that these tests again 
could have pretests and follow-ups, if important enough. 

6. Evaluation— to determine how successfully or unsuccessfully the pre- 
dictive findings worked out, and why. While it is too late to revise a piece 
of research after it is done, evaluation can contribute greatly to improved re- 
search in the future, and, indeed, can often uncover valuable information for 
current use. Thus, when one new product research project was evaluated three 
months after market introduction, it was discovered that a few retailers were 
selling the product at a slight reduction from the planned price— and getting 
almost twice the volume. 

In fact, there is much to be said, not just for a one-time evaluation, but 
periodic examinations spaced farther and farther apart, throughout the life 
of the product. And the more successful the product, the more need to guard 
against taking for granted that the way it is being marketed is in all respects 
the best way. 

Actually, the above six-step plan could be refined still further, by doing the 
main test in a number of stages— each one getting into the problem a little 
deeper, and more elaborately. A good example of such cumulative stages is the 
Fisher Company's research described in Case 2-3. While the total expenditure of 
time and money may be greater this way, the research— and the work on the new 
product, if that is what is being tested— can be shut off more quickly, or necessary 
changes made, if it looks as if the project is on the wrong track. 

The same reasoning indicates the desirability of early marketing research, no 
matter how rough, in the case of products involving expensive technical develop- 
ment—perhaps even before samples or models are available. Not only may such 
research come in time to save some financial commitment to a vain or unprofit- 
able venture, but it may help to avoid the build-up of a costly psychological com- 
mitment to going on with a project just because so much time, effort, and money 
have already been spent on it. 

All of this calls for a strong marketing research man, or director of marketing 
research in a larger company, who is considered a part of the top management 
team or who has the respect and confidence of the top executives. 



ELEMENTS OF THE TEST 

Now let us look specifically at the role of planning in the main test itself. To 
begin with, note how we have been continually using the word "test." It is a good 
idea to always think of research as testing— even if it's only testing the waters to 
find out if there are any fish there to be caught. Testing implies that you have some- 



PRODUCT POTENTIAL 141 

thing definite to test (the proposition, of course); that you have some test to put 
it to (which would be comparison); and above all, that you have a purposeful 
design to get useful results. 

That is usually the right order, too. Thus, the original need for marketing 
action suggests the proposition; knowing that leads directly to a search for the 
kind of comparison which will be most effective in throwing light on the proposi- 
tion; and then what is needed is a working plan to bring about the comparison in 
a form in which it can be observed and, at least roughly, measured. 

Such a working plan must cover these elements: 

1. The source of the information— usually a sample of people, with the 
critical issues being which people and how many of them to question or 
observe. 

2. The media by which to reach the source of the information— mail, tele- 
phone, personal interview, group discussion, observation. 

3. The instrument by which to extract the information— the question- 
naire, the interview outline, the test arrangements, and the observation 
devices. 

Weaknesses in any one of these three elements can spoil a test; and, un- 
happily, a spoiled test may be worse than no test at all, for the simple reason that 
it gives rise to a spirit of false confidence. So all three are important. But if there 
is ever need to grade them in terms of where you are more likely to make crucial 
mistakes, the first one is most important— particularly for the kinds of problems 
being discussed in this section; namely, product potential. 

The question of instrument becomes more pressing in connection with the 
more subtle nuances of competition and brand preference (which will be tackled 
in Section Five, where that whole area is the subject under consideration). But one 
suggestion is relevant at this point: do not stint in asking for demographic (some- 
times called "classification") data— age, education, sex, occupation, income class, 
and so forth. The reason is, that often, after the planned-for test information is 
collected, you find you could have made some meaningful comparisons if only 
you had secured more data to classify the different kinds of people who responded 
to your test. 



Selection of Test Media 

Let us take up the question of media before facing the more important and 
intricate problem of source of information (sampling). One immediate and over- 
all observation is called for; namely, that it is impossible to generalize as to what 
are the right media for various kinds of tests. 

Of course, some types of tests dictate their own media. For example: 

The rate at which people blink their eyes as they walk up and down the 

aisles of a supermarket is supposed to indicate the degree of interest they 



142 PRODUCT POTENTIAL 

have in the products they pass. Obviously, this requires either personal observa- 
tion or something like the use of a moving picture camera. 

To learn what programs people are following on TV or radio, one must 
either install some kind of recording device or use the telephone for spot 
checking of what programs they have turned on at the time of the call. (It is 
dangerous to ask questions afterward because people's memories are very un- 
reliable on something like this.) 

Keeping track of people's purchases over a period of time may call for 
use of a panel of consumers (either permanent, like the famous Chicago 
Tribune panel, described in the appendix to Case 5-2, or arranged for the 
purpose), equipped with diaries and other forms, and instructed in how to 
keep current records. 

Where other people are interposed between the tester and those he 
wants to put to the test— for example, executives with secretaries who screen 
their mail, and their personal callers too— sometimes telephone calls will get 
through. (The alternative is to make the importance of the research so evident 
that the secretary will clear the way.) 

When it is important to have actual corroboration of people's claims, 
then personal contact is called for, so the interviewer can, for instance, offer 
to supply a new razor blade for the old one that the interviewee says he has 
been using, thus checking on whether he specified the right kind or not. 
But beyond such clear-cut situations, the decision— which of the major media 
to use— will rest, in each instance, on a balance among a number of factors: 

(a) Cost. For a given number of opinions or other reactions, mail is 
usually cheapest, telephone next, the personal and small-group media most 
expensive. (Large meetings or conventions can be economical— but some- 
times unrepresentative.) 

(b) Need for depth of response (versus large numbers for projection; see 
discussion of sampling). Where people's reactions have to be watched or 
people have to be led to respond more freely and in more detail, the order 
is the reverse of that for cost. 

(c) Sensitivity of subject area. Thus, information of a personal nature 
(such as hygiene habits) is more likely to be divulged in proportion to the 
impersonality of the test, with mail first, telephone second, and the person-to- 
person situation last, again. One exception is that in a skillfully led group 
discussion, if the "ice can be broken" by one person speaking out on a deli- 
cate subject, the others will follow (in a snowballing effect), and the end 
effect is a freeing of inhibitions; but the appropriate caveat here is that this 
is both the most difficult to conduct and the most difficult to interpret after- 
ward of all means of securing information, and therefore should be expensive 
because it requires a rare degree of talent. 

(d) Need for honesty (versus people's desires to "look good," either to 



PRODUCT POTENTIAL 143 

the tester or to themselves). Depending on the particular situation, people 
will exaggerate their social, cultural, and personal accomplishments (including, 
in the case of women, their youth), or will understate their deficiencies in 
these areas. This is not deliberate falsification, but just natural "human na- 
ture," including the wish to be nice to whoever is conducting the test. The 
fact is simply that it is easy for a person to stretch the truth : to remember the 
one copy of The Atlantic in which he happened to read an article in a library 
when asked what magazines he reads, and to forget the four or five copies 
of Confidential that he read from cover to cover while waiting in the barber 
shop. A mass survey several years ago did, in fact, come up with The Atlantic 
showing ten times the readership of Confidential, though at the time their 
circulations were exactly the opposite. This is the kind of human contin- 
gency that must be carefully guarded against in planning the whole design of 
the test (including the sample and the information-getting instrument). 

In general, if the subject of investigation is something people would 
want to look good at in the presence of others, or spontaneously, the per- 
sonal interview will produce overstatement; whereas, if it is something people 
would want to look good at in their own minds, then the mailed question- 
naire will produce overstatement, because they have more time to think about 
it as they fill out their answers to the questions. 

The point is that the researcher must think of how people are likely to react 
in each instance, and then choose his medium in terms of the best balance of 
interpersonal requirements, information needs, and cost in that situation. 



SOURCE OF INFORMATION 

The question of media leads us right back to the question of people. Who 
shall be the source of information? This is a major question deserving major con- 
sideration. 

More specifically, what kind of people shall be our source of information, and 
how many do we need to put to the test? While there is good reason to believe 
that the identity of the people may be more important than the number, the latter 
question usually gets more actual attention because it lends itself to techniques 
that are reassuringly definite and generally applicable. On the other hand, because 
getting the right people for the sample is a reflection of the whole purpose of the 
test, it varies from situation to situation, but that does not lessen its importance. 

Kind of People 

More misleading findings can result from this factor being askew than from 
not having a large enough sample. The classic example of such a failure is the 
debacle of the Literary Digest poll in the Presidential election of 1936, which 



144 PRODUCT POTENTIAL 

brought such ridicule on that then-flourishing magazine that it went out of busi- 
ness shortly thereafter. The sampling lists were telephone directories and lists of 
automobile registrations. Deficiencies resulted at two levels: 

(1) The sampling lists favored the higher-income group; those who 
couldn't afford telephones or cars weren't even represented in the lists, and 
they were a larger part of the total population in those days. 

(2) Of the people on the lists who received questionnaires, a higher pro- 
portion of those who were better educated (and also of higher income) filled 
out and returned the questionnaires than did those on the other end of the 
scale. 

Such a distortion had never bothered the Literary Digest polls before, but in 
this election the voting followed a new economic pattern— and the prediction of 
victory for Landon was not only wrong, but overwhelmingly wrong. As a matter 
of history, Roosevelt won by a landslide. 

The problem is not just the original question of what people constitute the 
particular "universe" that should be tested, or what list can be used to represent 
that universe for sampling purposes, but, also, who out of that list will send in 
their questionnaires, or be at home and willing to be interviewed, or happen to get 
observed— that is, the "effective sample." Note the important distinction between: 
(a) the universe being tested; (b) the sampling list, made up of the kind of 
people or phenomena that are considered to represent the universe; and (c) the 
effective sample, that is, the people or phenomena that actually get covered in 
the final results. 

SAMPLING LIST. Today, telephone directories are regarded as a good 
sampling list for a universe of the general population, since a much larger pro- 
portion of the total population (almost 80 per cent) now have telephones and for 
most purposes the people with telephones account for just about all of the mass 
market that amounts to anything. For some really "high-ticket" products (for ex- 
ample, a genuine pearl necklace) there is a worse defect at the other end of 
the economic and social scale: in metropolitan areas like New York, many 
of the most wealthy and prominent people have unlisted numbers and aren't in 
the directory at all (about 2 per cent of total telephones are in this category). 

On the other hand, many lists or locations, particularly if they depend on the 
voluntary action of people, are likely to represent the "cream of the crop" and 
therefore to be unrepresentative of their particular universe. Thus, motel owners 
who subscribe to a trade journal in that industry are likely to be the more suc- 
cessful, larger operators, and therefore are not representative of the whole universe 
of motel owners; and people who live in certain areas or shop at certain stores 
may be particular kinds of people, and to that extent are not representative of 
the universe of buyers of this or that product. 

For this reason, it is usually better to use lists prepared by others— for ex- 
ample, census lists such as many police departments keep; compiled lists of people 
by income, occupation, and so forth, such as mailing services or other agencies 



PRODUCT POTENTIAL 145 

maintain; and, again, telephone directories, which have the inestimable advantage 
of being scrupulously accurate and up-to-date. Similarly, in the event of going to 
places to interview or observe people, it is better to use a map or list of dwelling 
houses or compilation of stores that contains all elements of the universe in 
question. 

If the universe itself is selective— for example, if a product to be tested is 
designed only for certain kinds of people, then a selective list is appropriate. 
Also, if such a list exists, it is likely to be relatively more efficient for sampling pur- 
poses (both in terms of representativeness and in terms of getting response) than 
a list used to find out something about "the average man" or a cross-section of the 
total population. 

EFFECTIVE SAMPLE. The actual returns, interviews, or observations that 
one succeeds in getting from the universe represented by the sampling list is what 
counts. 

Thus, if there is anything about the test that is likely to cause certain kinds 
of people to enter the effective sample disproportionately, then one has to be on 
guard against this distorting of the results. In general, better educated people are 
more likely to be alert and cooperative, so if the product test involves a matter 
of literary or current knowledge (including awareness of advertising), then the 
reactions of that segment of the population will figure more heavily in the results 
than should be the case on the basis of actual numbers— and this may or may not 
be serious, depending on whether that segment is likely to have a different set of 
reactions than the rest of the population. In the case of a survey of irradiated 
foods (Case 3-2), for example, the less literate people interviewed turned out to 
have a definitely less enlightened attitude. 

A carefully thought-through proposition is the best safeguard against first- 
level and second-level distortion (distortion through selecting, first, the wrong 
universe or, second, the wrong sampling list or location to represent it). But what 
about third-level distortion— disproportionate numbers of certain kinds of people 
or phenomena out of the sampling list or location coming into the effective 
sample? (Fourth-level distortion, resulting from the make-up of the information- 
getting instrument will be treated in Section Five.) 

This third-level distortion is a real danger, and a high percentage of returns, 
responses or observations is not necessarily any safeguard. In fact sometimes the 
very factor that has a distorting effect itself causes a high percentage of people to 
cooperate or show up in the test. Take these two situations. 

( 1 ) Seventy per cent of the people know but are not very interested in the 
product or what it does, so when a questionnaire asking people if they know 
it is sent out to a random sample, 8 per cent of these people and 8 per cent 
of the balance (who do not know it) fill out and return their questionnaires; 
the results of the effective sample correctly show 70 per cent with a knowl- 
edge of the product: (.08 X .70) -^ [(.08 X -70) + (.08 X 30)] = .70. 

(2) Thirty per cent of the people know and are very interested in the 



146 PRODUCT POTENTIAL 

product, so 40 per cent of these people return their questionnaires, and again 8 
per cent of the balance do so; and the results seem to show, very misleadingly, 
that 68 per cent of the universe know the product: (.40 X .30) -f- [(.40 X 
.30) + (.08 X -70)] = .68. Yet all those who returned their questionnaires 
answered correctly. 

In general, the distorting effect is likely to be caused by a larger than true 
proportion of responses or reactions from: 

People who are proud of their own records in the area of the test. 

People who are particularly interested in the subject of the test. (A good 
accompanying letter of explanation can help to spread this interest, thus 
resulting in an effective sample that is not only larger, but better, in the 
sense of being more representative.) 

To illustrate: when a business publication sent out a short, succinct postcard 
questionnaire, and then a follow-up with a twenty-five cent coin attached to it, 
a total of 65 per cent returns was received. Another business publication sent out 
a complicated (but challenging) four-page questionnaire to exactly the same list 
and got a 15 per cent total return. Yet the latter mailing actually secured a more 
representative return in terms of the actual proportions of top- and middle-range 
executives. What happened was that the first one got through to the greater 
number of middle managers better than it did to the smaller number of presidents 
and vice presidents, whereas this was not the case with the latter, because it had 
been made more relevant to top management problems and interests. 

This is not just a problem of mail questionnaires. Certain kinds of people will 
allow themselves to be interviewed more readily; or will respond more enthusi- 
astically to the test, whatever it is; or will just be the kind of people that are 
likely to be in the store or wherever else they are being observed. 

Also, it is worth noting that the problem is not basically different in foreign 
countries. People are people the world over, and one of the basic attributes of 
human nature is friendliness. Where questionnaires and interviews are less well 
known, there may be need of more patience, more preparation, and more skill in 
interviewing, but experience indicates that the handicaps can be overcome. All 
the national stereotypes are being disproved— like 7 "An Englishman's home is his 
castle and he won't allow an interviewer in," or, "A Frenchman won't even tell 
his wife about his business, let alone a stranger," or, "A Mexican will put a knife 
in your back if you knock on his door at the wrong time of day." The only 
problem in some countries is lack of physical facilities in the form of mailing 
services, census lists, trained interviewers, or just sheer illiteracy— and thus the only 
problem is one of ingenuity on the part of the researcher. 

No, abroad as in this country, the problem of third-level distortion is likely 
to be one of too much cooperation (on the part of certain people), not too little. 

GUARDING AGAINST DISTORTION. In this book, the term "distor- 
tion" is used to cover all phenomena, except random error, which cause findings to 
be unrepresentative of the universe that the researcher desires to test. The technical 



PRODUCT POTENTIAL 147 

term "bias" is avoided entirely because it popularly connotes willful slanting of 
opinion or shading of facts, whereas the phenomena we are concerned with result 
from spontaneous reaction or unintentional misjudgment. To meet this kind of 
problem, there is only one real safeguard: actual follow-up to check whether those 
who did not come in to the effective sample are the same in all essential char- 
acteristics, and would have responded the same except for inertia or happenstance. 

It does help to compare the effective sample against the original list or loca- 
tion where this is possible (in terms of age, position, education, length of product 
use, and any particular factor likely to affect the role of response). It also is com- 
forting, but not necessarily conclusive, if succeeding mailings to the same list of 
people continue to bring in the same findings. But the trouble is that one doesn't 
always know what the crucial factors are, or whether there still isn't a nonrespond- 
ing residue that would react entirely differently. Unless those who do not get into 
the effective sample are checked on the subject of the test itself, distortion can 
result. 

On personal or telephone interviews, the researcher can keep calling back 
(keeping a separate record of call-back results, to see if they are any different) until 
he satisfies himself that he has "closed the gap." With mail questionnaires, the 
problem is more difficult. However, researchers have found it useful to code ques- 
tionnaires against the mailing list so as to be able to identify nonrespondents, and 
then to make a telephone follow-up, keeping after a random sample of, say, 100 
nonrespondents until either a nondistorting reason for failure to "come through" is 
uncovered (for example, vacation, death, removal to a new address, and so forth), 
or they are actually located and queried on a few key questions. 

Size of Sample 

Such follow-up becomes particularly important when one has carefully cal- 
culated the size of sample so as to be able to project the findings from the sample 
to the universe with specific confidence limits; for instance, to be able to say that 
since 67 per cent of the people in the sample (the effective sample, it must be) 
like this product, we can be 95 per cent confident that from 65 per cent to 69 per 
cent of the people in the actual universe like the product. Such projections are 
based on the size of the sample, but mean absolutely nothing unless there is rea- 
son to believe that the quality of the sample also is good. 

The procedures of "probability" sampling are designed to produce an ef- 
fective sample which is representative— to produce it this way just because, by 
design, the operations of chance are unimpeded or unaffected by human judgment 
on the part of the researcher in selecting the particular units of the universe to be 
in the sample for testing. But it still is wise to see, afterward, whether chance did, 
in fact, operate evenly. As we have just seen, the human element in the sample 
itself can distort the responses or reactions. Also, the very fact that we are relying 
on chance means that every possible combination of events can happen— that 
someone can throw a seven with the dice 10 or 30 or 100 times in a row, or that a 
monkey banging on the keys of a typewriter can write Hamlet. 



148 PRODUCT POTENTIAL 

Fortunately, the more "chance" you give chance (that is the more oppor- 
tunities you give it), the more likely it is to work in your favor. It is true that on 
any single toss of a penny, even if ten heads have preceded, the odds are still 
exactly the same 50-50 that it will be a head or tail on the next toss. One cannot 
count on that toss beginning to redress the balance. Yet it is also true that the 
more times one tosses the penny, the more likely it is that the sum total of tosses 
will come out 50-50 heads and tails. This is the Law of Large Numbers (there is 
no Law of Averages), and for us, here, it means that the larger the sample, the 
greater the probability that our findings will be closer to the true figures for 
the actual universe. And we can put a figure on that probability in each instance. 

CONFIDENCE LIMITS. Think of it like sampling a deck of cards to find 
out what per cent is red. If you draw one card, there are two possibilities— it can 
be either red or black, which makes the universe look like 100 per cent red or 
per cent red, and the closest you can get to the truth is 50 percentage points 
away. But if you draw, say, three cards, there are eight possibilities: 

1. First card red, second card red, third card red— 100 per cent red 

2. First card red, second card red, third card black— 66 2 A per cent red 

3. First card red, second card black, third card red— 66 2 A per cent red 

4. First card red, second card black, third card black— 33V3 per cent red 

5. First card black, second card red, third card red— 66 2 A per cent red 

6. First card black, second card red, third card black— 33 l A per cent red 

7. First card black, second card black, third card red— 33 1 /3 per cent red 

8. First card black, second card black, third card black— per cent red 
These add up to: 

No. of Possibilities Combinations Percentage Red 

1 3 red 100% 

3 2 red, 1 black 66% 

3 1 red, 2 black 33 l / 3 

1 3 black 
8 

And 6 out of 8 of these possibilities are within 16 2 A percentage points of the 
correct universe figure of 50 per cent. (While this reflects the working of the Law 
of Large Numbers, the actual number in a sample, as we shall see in more detail 
shortly, does not have to be as large as you might expect to get useful results.) 
Similarly, with a draw of 5 cards, there are 32 possibilities, as follows: 

No. of Possibilities Combinations Percentage Red 

1 5 red 100% 

5 4 red, 1 black 80 

10 3 red, 2 black 60 

10 2 red, 3 black 40 

5 1 red, 4 black 20 

_1 5 black 

32 



PRODUCT POTENTIAL 



149 



Note that here there are 20 chances out of 32 (63 per cent) that the sample will 
produce a finding within 10 percentage points of the actual universe figure, and 
another 10 chances (making a total of 94 per cent) that the sample finding will 
be within another 20 percentage points. 

If all these possibilities were piled up like so many coins, they would look 
like this: 




100% 90 80 70 60 50 40 30 20 10 



The larger our drawing from the universe, the more pronounced would be 
the shape; that is, the larger would be the proportion of coins piled up in the 
middle; or, in terms of sampling, the greater the probability that the sample find- 
ing would be closer to the actual universe figure. What we have is the so-called 
normal bell-shaped curve, which (for samples of 30 or more) can be generalized 
like this: 




99.7%. 



In statistical terms, 68 per cent of all possible drawings from the universe will 
be in the indicated middle portion of the area under the curve, 95 per cent will em- 
brace that and the additional area flanking the middle, and 99.7 per cent will 
of course cover almost all the total area. (Actually, these proportions apply only to 
a large sample from a large universe; special formulas must be used if the sample 
is less than 30 or if it amounts to more than 5 per cent of the universe it is drawn 
from.) 

Lines b and c represent "one standard deviation" from a; d and e represent 
"two standard deviations"; and f and g "three standard deviations." (Actually, for 
some purposes, the term "standard error," usually represented by the Greek sigma, 
C is more precise, but it is an unnecessary distinction here.) 

The curve will change in altitude and width, becoming higher and narrower 
with an increase in the size of the sample, but the bell-shape character will remain 
(assuming, as we need to do in sampling, that deviations from the true universe 
figure are evenly distributed in the universe itself). So the actual size of a standard 



150 PRODUCT POTENTIAL 

deviation will differ, but it will still cover the same proportion of total sample 
possibilities. Expressed in a formula: 

q 



=#= 



where S is a standard deviation; n is the number of drawings from the universe, 
or the number in the sample; p is the percentage sample finding (for example, 60 
per cent red in the case of the deck of cards illustration ) ; and q is the complement 
of p (that is, if p is 60 per cent, then q is 40 per cent; or if p is 70 per cent, then 
q is 30 per cent). This is how it works: 



s 


_ A /.80 • .20 
v 100 


s 
s 

and 2S 
and 3S 


= \.0'016 

= .04 
= .08 
= .12 



That is, we can be 68 per cent confident (S) that the finding from our sample 
(80 per cent) is right within four percentage points plus or minus; that is, 
that the true universe figure is from 76 per cent to 84 per cent. (Literally, "the 
probabilities are that 68 times out of a hundred a sample of this size will produce 
a finding this close to the true universe figure.") Or, we can be 95 per cent con- 
fident (2S) that we are within 8 percentage points of the true figure; or 99.7 per 
cent confident (3S)— virtually certain— that we are within 12 percentage points. 
For most marketing purposes, 95 per cent confidence is all that is needed; or, to 
put it another way, the extra confidence of 99.7 per cent is not worth the extra 
cost of the larger sample needed. 

If the sample were 400 rather than 100, then the computation would come 
out with S = .02, 2S = .04, 3S = .06. With the larger sample, we can have 
either the same degree of confidence but more precision (that is, greater closeness 
to the actual universe figure) or a greater degree of confidence and the same pre- 
cision. 

Notice, also, that if our finding were 50 per cent— that is, if 50 per cent of 
the sample said they liked green cheese— then our formula would be: 



-<> 



50 • .50 



100 

S =\.0025 
S = .05 
and 2S = .10 
and 3S = .15 
This gives us less confidence and /or less precision — which is only natural, since, 
if 50 per cent feel this way, the opinion is nowhere nearly so clear-cut and so 
prominent in the universe. And we can be only 95 per cent confident that the true 
figure of green-cheese likers is somewhere between 40 per cent and 60 per cent. 



PRODUCT POTENTIAL 151 

For an arithmetic "mean," or average, rather than a percentage, the corre- 
sponding formula is: p. 

where D is the degree of dispersion in the sample (sometimes called the standard 
deviation of the sample, which is different from the standard deviation of the 
finding). D is similar in its effect to the p • q in the percentage formula, and is 

actually a shorthand way of saying \j — _ — with 2 meaning "sum of" and "x" 

* n — 1 

standing for how far from the mean each individual unit is. Thus, if we had a 

sample of 17 young men from a large club and wanted to know the average, or 

"mean," age of club members: 







Distance from 






Age 




Mean 


Squared 


1 


23 yrs. 


20 


- 23 = -3 


9 


2 


20 


20 


- 20 = 





3 


17 


20 


- 17 = 3 


9 


4 


20 


20 


- 20 = 





5 


22 


20 


- 22 = -2 


4 


6 


24 


20 


- 24 = -4 


16 


7 


20 


20 


- 20 = 





8 


20 


20 


- 20 = 





9 


22 


20 


- 22 = -2 


4 


10 


17 


20 


- 17 = 3 


9 


11 


20 


20 


- 20 = 





12 


18 


20 


- 18 = 2 


4 


13 


18 


20 


- 18 = 2 


4 


14 


20 


20 


- 20 = 





15 


18 


20 


- 18 = 2 


4 


16 


20 


20 


- 20 = 





17 


21 


20 


- 21 = -1 


1 


am 


340 yrs.- 


r 17 = 20 


Sx 2 - (n- 


D = 64-r ( 



The D is therefore "^4^ or 2; and applying this to our formula 

s - w 

S = 2 -T- 4.1 

S = 0.5 
and 2S = 1.0 
and 3S = 1.5 

Which means, taking the 2S, that 95 times out of 100 a sample of this size will 
produce a finding that is within one year of the true figure for its universe, or, in 
management terms, we can be 95 per cent confident that the true figure is be- 
tween 19 and 21 years. 

There are some short-cut ways of figuring D for a sample mean, and it can 
always be done on a computer. For most purposes, however, D can be estimated 



152 PRODUCT POTENTIAL 

by visual inspection. For example, looking over the above figures, or 10 or 100 times 
as many figures of the same general kind, one can see fairly easily that 2 will cover 
all but a small amount of the dispersion; or if there are very many units in the 
sample, then a sample of the sample can be taken for purposes of estimating D. 

In either case— percentage or mean— S usually depends more on n than it 
does on the degree of dispersion (^p • q or D), if only because of the fact 
that the researcher can increase n at will. And one can always hedge— that is, be 
sure of having more than enough in the sample to get the desired confidence 
and/or precision, or be sure of understating the confidence and/or precision in a 
given sample size. 

Some conservatism is already the effect of entering into the formula the 
degree of dispersion in the sample; it really should be the degree of dispersion in 
the universe, but the sample dispersion is usually the only practically available 
figure— and it is likely to be about the same, or a little on the high side, simply 
because there aren't as many numbers to wash out erratic differences. But to be 
doubly sure, one can always use the maximum p • q which is .50 • .50, or at least 
something nearer such an evenly divided amount, that is, such a vague and un- 
pronounced state of opinion; or use for D a figure which covers all or virtually all 
of the differences between the mean and the unit figures, no matter how extreme, 
which would have the same effect as if, again, the trend in the universe were vague 
and unpronounced. 

In any event, note that the formula can be used two ways: (a) in advance 
of the test, that is, solving for n to see how big a sample is desired; or (b) after 
the sample results are in, solving for S to see how much confidence and/or preci- 
sion should be accorded the results. Above, it has been illustrated the latter way. 
But it is just as practical to use it the former way, or for that matter, both 
ways. Thus, suppose management indicates that the marketing decision, to be 
made on the basis of the test, is important, which the researcher translates into 
''Management should be able to have 95 per cent confidence that the sample find- 
ing is within four percentage points of the universe figure." Assuming that the 
finding is expected to come out somewhere around 60 per cent (so p • q = 
.60 • .40), this would be expressed mathematically: 

2S = .04 
S = .02 



Then, .02 = J - 60 ' A0 
* n 



.0004n = .24 
n = 600 

So we plan on a sample of 600. Then, after the results are in, p can be 
checked to make sure that it is in fact not less than 60 per cent and, therefore, 
that the desired precision has been secured. Also, if, to get an effective sample of 
600, we had to send out 3,000 questionnaires or had to undertake 750 interviews, 



PRODUCT POTENTIAL 153 

then the actual n of the effective sample can be checked and the confidence and/or 
precision refigured. Remember, it is the effective sample that counts. 

Further, if one wants to draw conclusions about a subsection of the sample, 
then that part must have the requisite number. Thus, if we expected 10 per cent 
of our sample to be over 50 years old, and we wanted to know, with 95 per cent 
confidence and precision of ± four percentage points, how people of that age felt 
about our new proposed product, then we would have to have an effective sample 
of 600 for them. This would call for an effective sample over-all of 6,000; and if 
we were sending out mail questionnaires and expected a return of 20 per cent, we 
would have to have a mailing of 30,000! And then, when returns came in and 
were classified by age, or, if, as we proceeded in our analysis, we saw a new sub- 
section that we wanted to draw conclusions on, we would have to refigure again 
to see just what confidence and /or precision we did achieve. 

A further use of the confidence-limit formula is in appraising the relation 
between two different findings in the same sampling (say, percentage of interest 
in Product A and percentage of interest in Product B) or between similar find- 
ings in different samples (say, percentage of interest in Product A this year and 
then again next year). However, that will be discussed in Section Five. 

COMMON MISUNDERSTANDINGS. Only too often, people in and out 
of business fail to understand two things about probability sampling: 

(1) It makes virtually no difference how large a proportion of the uni- 
verse a sample is. As far as size and confidence limits are concerned, a 
sample of 100 is just as good for a universe of 10,000,000 as it is for a uni- 
verse of 10,000. This is hard for anyone to accept since it seems to run against 
common sense, but the explanation is that, if phenomena in the universe are 
evenly distributed, a drawing of 100 is just as likely to cover the range of opin- 
ion and reaction in the case of 10,000,000 as in the case of 10,000. It is the 
number drawn for the sample that determines how well the variations wash 
out; 100 tosses of the coin will tend to wash out runs of heads or tails, and 
come out close to 50-50, no matter how many coins there are in the world 
and how many people there are to toss them for how many times. (Of 
course, a universe of 1,000,000 may be a more important market, and there- 
fore more confidence and/or precision may be desired, but that is a matter 
of weighing costs against usefulness, rather than figuring probabilities.) 

(2) A sample of 100 or 500 may produce all the confidence and/or preci- 
sion that the marketing decision in question justifies. After all, many market- 
ing decisions are concerned with fine differences between alternatives, where 
either alternative will be reasonably satisfactory. If the color of the ink on the 
label is only a small fraction of what determines whether people will buy 
the product, either blue or green won't have much of an effect, though one 
may be somewhat less desirable; in such a case almost any sample size is 
enough even if the chances are only 51 out of 100 that it will yield the right 



154 PRODUCT POTENTIAL 

answer. Or take a situation where it will be profitable to go ahead with a 
product whether the share of market is 40 per cent or 80 per cent; in such 
a case, what need is there for precision, or for a sample made large just to 
increase precision? 

But what happens oftefr is that the sample is made excessively large just to 
impress top management. This is not to deprecate the usefulness of figuring sample 
size and confidence limits. Far from it. In fact, an understanding of these matters 
may pay because it helps to avoid unnecessary expenditures on marketing research 
for routine decisions, and also because it calls for a larger sample when that fits 
the truly crucial decision that keeps coming up from time to time. (On the low 
side, however, one useful rule of thumb is that any problem that is worth drawing 
a sample for at all is worth a sample of 100— that being the minimum figure that 
takes care of most sampling contingencies.) 

Sample Design 

But the value of probability sampling goes farther than actual projectability. 
The essence of such sampling is that you know in advance that every unit in the 
particular universe has an equal chance of being drawn in the sample. And the 
only way to do this is to so design the sampling plan that, once it starts, no human 
judgment can affect its course. Probability sampling is sometimes called random 
sampling, because, looked at in another way, it is designed to leave random 
chance free to operate, but the term is often interpreted as meaning haphazard, 
which it definitely is not. In fact, only the most careful planning can avoid 
distortions like this: 

(a) An interviewer wishing to get a good sample of women shopping 
in a supermarket decides to question women coming out of the store on a 
random basis, but without realizing it, he instinctively walks up to the younger, 
prettier women; or the home interviewer instinctively selects the cleaner, more 
pleasant-looking homes to call on. What is needed is a plan (say, every tenth 
one, regardless of appearance) that keeps the human trait from introducing 
judgment and impeding true chance. 

(b) A researcher wants to sample the buying habits of customers of a 
department store, where 60 per cent of purchases are made for cash, and 40 
per cent are charged. To be sure to have both kinds represented, he takes 
100 cash transactions and 100 credit transactions— thus giving each credit cus- 
tomer a^-rth chance of being drawn in the sample, and each charge cus- 

40 
tomer only a-^:th chance. What the researcher must do is to mingle all the 

cash and credit transactions together and draw 200 out at random. 

Then, if they are all numbered or can be numbered easily, for example, he 



PRODUCT POTENTIAL 155 

can make a published table of random numbers (with row after row of num- 
bers arranged by chance) and, starting somewhere in the table at chance, fol- 
low a definite plan, like two down and three to the right, until he has the 
requisite sample; this is a routine that is equivalent to drawing the names 
blindly out of a hat. 

Or he can shuffle them all together so their arrangement is random, and 
then take every tenth one, or twentieth, or whatever it is that is needed to 
produce the requisite sample. (Incidentally, the names in a telephone book 
are usually regarded as a random arrangement, on the ground that alphabetic 
order is arbitrary, and independent of marketing factors; whereas a chronolog- 
ical listing of department store customers would definitely not be "random," 
because time does reflect daily, weekly, monthly, and seasonal patterns of 
store patronage.) 

Or, as described in the section immediately following, our researcher 
could solve the particular problem by stratifying his sample. So far, we have 
discussed sampling as applied to the total universe that is being tested — simple 
probability sampling. There are two major variations of probability sampling 
worth keeping in mind: stratified sampling and cluster sampling. 
STRATIFIED SAMPLING. Where there is any reason to believe that the 
universe is made up of strata which are internally homogeneous and externally 
quite dissimilar (as might be the case with the department store customers above 
—that is, the charge customers quite alike, the cash customers quite alike, but 
the two groups quite different from each other), then it is statistically more effi- 
cient to use stratified sampling. 

The confidence and/or precision will be greater for the same size sample. 
Though we shall not go into the details of computation here, in effect S is a 
weighted average of the S's for all the strata, but those individual S's are smaller 
(that is, indicating greater precision) because dispersion inside each stratum is less. 
The need and opportunity for stratified sampling is great in marketing, be- 
cause in so many marketing situations the universe is askew. In almost any com- 
pany, 20 per cent or 30 per cent of all customers account for 70 per cent or 80 
per cent of all business. Or take a situation where the researcher is trying to find 
the average inventory of his product in stores, and the actual universe breaks down 
this way: 



Number 
of Stores 


Size of 
Store 


Inventory 
per Store 


Total 
Inventory 


10,000 

90,000 

900,000 

1,000,000 


Large 

Medium 

Small 


500 cases 

40 cases 

3 cases 


5,000,000 cases 

3,600,000 cases 

2,700,000 cases 

11,300,000 cases 



Average (11,300,000 t 1,000,000) = 11.3 cases 



156 



PRODUCT POTENTIAL 



Now suppose we decide to draw a sample of 200, and everything works per- 
fectly, so we have exactly 1/ 5,000th of each stratum: 



Number 
of Stores 

2 

18 
180 
200 



Size of 
Store 

Large 

Medium 

Small 



Inventory 
per Store 

500 cases 

40 cases 

3 cases 

Average (2,260 r 200) 



Total 
Inventory 

1,000 cases 
720 cases 
540 cases 

2,260 cases 

= 11.3 cases 



But if the "luck of the draw" misses just one large store, and brings in a 
small store instead: 



Number 


Size of 


Inventory 


Total 


of Stores 


Store 


per Store 


Inventory 


1 


Large 


500 cases 


500 cases 


18 


Medium 


40 cases 


720 cases 


181 


Small 


3 cases 


543 cases 


200 






1,763 cases 



Average (1,763 * 200) = 8.8 cases 



Having reason to believe, or just suspecting, that there is this much difference 
between different-size stores, the researcher can stratify his sample so as to get a 
proportionate sample from each stratum: drawing two at random from the list 
of large stores, 18 at random from the medium-size stores, and 180 at random from 
the small stores. Or he can go ahead with his total sample, regardless, and then 
weight his results according to the size of the strata in the universe— a kind of 
stratification after the event, but inefficient, and indeed unnecessary if he has the 
strata lists and knows their size in advance. Then he is deliberately running the 
risk of getting less of the large stores or important customers and having less 
precise figures for those strata when he weights them up in the final result. It 
makes more sense to oversample in the important strata purposefully (that is, by 
designing the sample this way in advance), and then to weight down those strata 
to their proper proportion in the final result. 

It can be stated, categorically, that if stratification can be done with little 
expense or bother, it should be. For the same reason it often should be done even 
when it requires considerable expense and bother. The reason is that the gain in 
efficiency is enough to justify the extra effort and cost. 

CLUSTER SAMPLING. When doing large-scale "area" sampling (that is, 
selecting locations to go to for interviews), it becomes very expensive, in fact 
prohibitive, to send interviewers to every house or store or other unit that is drawn 
in a simple random sample. Think of a sample of 500 spread over the United 
States, and the time and travel that would be incurred. Even for 100 or 200 



PRODUCT POTENTIAL 157 

spread across a large city, it would be laborious to go to 20 Ash Street, 25 Sawyer 
Street, 96 Lincoln Way, and so on for the whole sample. So, recourse is made to 
cluster sampling, which (as its name implies) is to draw clusters of names or 
addresses at random and then draw a number of units from each of those clusters 
at random. 

Cluster sampling is statistically less efficient than simple random sampling if 
the elements in each cluster tend to be similar, as usually is the case in area 
sampling. For example, the homes in a block or other area of a city tend to be 
like each other in their social and economic characteristics, hence also in their 
occupants' opinions, attitudes, and reactions to marketing moves. Thus, to take a 
number of sampling units from one block, rather than spreading them over that 
number of blocks, gives other, different opinions and information less chance to 
show up in the sample. 

In contrast, the loss of efficiency would be minor in using cluster sampling 
on a telephone book or other alphabetic directory, where it is assumed that the 
order of the names and the pages they happen to be on is in itself random, and 
has no other meaning or relationship to any marketing factor. Even the Smiths 
would all be different in economic, social, and even racial characteristics; included 
would be original Schmidts and Smythes, as well as plain Smiths. Here, the proce- 
dure would be to take, say, every 25th page, and then take four names from each of 
those pages— perhaps following a pattern like the fourth from the top and the 
third from the bottom on the left column, and the seventh from the top and 
the sixth from the bottom on the right column; or almost any pattern as long as it 
is decided on in advance. 

It should be emphasized, once again, that deciding in advance on some ar- 
bitrary design, like that just described for taking names from the pages of a tele- 
phone book, is important. Even sticking a pin into a page with your eyes shut 
can show subjective influence. Suppose you are using a map to determine what 
homes to call at. The pin-sticker is much more likely, unconsciously, to keep to- 
ward the center of the map, thus giving homes there a greater chance to be drawn 
in the sample than homes in outlying areas— which keeps it from being a true 
probability sample. 

Having a planned design is particularly important in area sampling, because 
it is so easy for the interviewer to let himself or herself be led unconsciously toward 
the more attractive houses. (It is usually herself, since women do better at inter- 
viewing housewives.) So the interviewer most often has definite addresses to call 
at; or she is given an arbitrary design, such as: go to the corner of Chestnut and 
Duke Streets, take the second house on the left from the corner, then the fourth 
house on the right, then turn left at the next corner and take the third house on 
the right, and so on; or she is instructed to count the houses in each block as she 
comes to it, and then to use a series of random numbers, like 269431857, taking the 
second house in the first block, the sixth in the second block, the ninth in the 
next block, and so on, repeating the series as often as she has to. 



158 PRODUCT POTENTIAL 

It is not surprising that a staff of field interviewers needs supervision. Very 
often those who do this work are on a part-time basis, including even men and 
women who have graduate degrees in social science, who may be teachers or house- 
wives themselves, for instance, and are "on call" with various marketing research 
agencies. In general, interviewers need very careful instructions to begin with; 
then their work has to be spot-checked to make sure they follow instructions — and, 
indeed, that they make the calls at all and don't just sit comfortably at home or 
in a hotel, imagining the interviews and filling out the forms accordingly. 

DESIGN OF AREA SAMPLE. As to the make-up of the area sample, it can 
vary from only a very simple design like four homes in each of 25 blocks in one 
city (a sample of 100) to the same design in each of 15 cities in each of 10 
states in each of 4 regions (a sample of 60,000). Or it can be a combination 
stratified-cluster sample, where, say, a proportionate number of cities of various 
size (or other characteristics) are drawn, and then cluster sampling is done within 
those cities. But in all cases there are three factors to be kept in mind: 

(a) It is more efficient to have more of the larger units, fewer of the 
smaller— for example, two homes in 50 blocks is likely to be more efficient 
(and more expensive) than four homes in 25 blocks. This is because of the 
tendency, mentioned above, for homes in the same block to be similar. 

(b) If interviews cannot be secured at the designated houses, it is im- 
portant to call back until they are— but this can be very expensive. In some 
surveys, interviewers are instructed to take the next house, but this means 
that any house next to a noninterview house has twice the chance of being 
drawn. This is not a serious defect, since the noninterview itself is a random 
occurrence and subjective judgment is not involved; however, it is better to 
have provided in advance for additional or spare units and to simply go on 
with the planned procedure. Or, there are mathematical techniques which, 
on the basis of pushing through interviews with a sample of original non- 
interviews, can roughly correct for the effect of all the noninterviews. 

(c) Each house (or store, or whatever it is) should have an equal chance 
to be drawn. Therefore, it is not right (and can seriously distort the results if 
economic or social status is involved) to draw the same number of houses 
from crowded blocks as from blocks with fewer, larger houses. Each unit in 
a crowded block has less chance to be drawn. Consider a block with 30 
houses, and one with 15, and suppose you are taking a sample of 24 out of 
100 blocks and four houses to each block, a total of 96. Then the chances of 
each unit in the two kinds of blocks would be: 

W X ^ = TI5 ( ° r 4 chances out of 125) 
i* X ^= -~ (or 8 chances out of 125) 

In such a situation, you must stratify on the basis of proportionate number 



PRODUCT POTENTIAL 159 

of dwelling units, or arrange that crowded blocks have twice as much chance of 
being drawn as the uncrowded blocks, so that (still with a sample of 96) : 

100 A 30 375 

100 A 15 375 

Of course— to be finicky— if one is interested in individuals rather than dwell- 
ing units, it would be necessary to make the same kind of provision for crowded 
houses (and apartment houses). 



QUOTA SAMPLING 

The exact opposite of probability sampling is quota sampling, sometimes called 
judgment sampling because the judgment of the researcher and/or his interviewers 
is used during the sampling procedure until the "quotas" are filled. Whereas prob- 
ability sampling should, by the unimpeded, unswayed action of chance, produce 
a representative or unbiased sample, quota sampling sets out deliberately to make 
the sample come out representative on certain predecided factors. 

Thus, if there are 52 per cent females and 48 per cent males in the population, 
questionnaires are sent to that proportion of females and males, or the interviewers 
or observers are instructed to get those quotas of the sexes. The same thing can 
be done with age or income class or city-size or whatever is deemed important. 
Complex, large-scale surveys can contain quotas of many different factors simul- 
taneously. 

DEFICIENCIES. Quota samples have a number of possible deficiencies com- 
pared with probability samples: 

( 1 ) The quotas have to be based on census or otherwise published figures 
—and these can be less than accurate, particularly through becoming out of 
date. 

(2) When the quotas have to be filled by the judgment of interviewers- 
say, quotas of women of certain age ranges, there is a tendency to have an 
undue concentration in the middle of the various quota ranges— more than a 
proportionate number of women around 25, 26, and 27 years in the 20-30 
range, for instance— just because of interviewers trying conscientiously to be 
sure that the women do belong in that range. Or conversely, in the event that 
the quotas are hard to fill, there is a temptation to carelessly take some that 
may be over the edge of the range limit— maybe some who are 19 or 31 years 
old. 

(3) One of the purposes of the quota system is to ensure that the sig- 
nificant factors are represented proportionately. But what if you don't know all 
the factors that are significant? Suppose you are making a test to find out 
something about people's attitude toward playing the tuba. Presumably, you 



160 PRODUCT POTENTIAL 

are very careful to get the right proportions of age, sex, education, and so on. 
What you don't realize, however, is that families of German descent are the 
mainstays of the tuba market (this is just a supposition for purpose of illustra- 
tion), and in your effort to get the other factors represented, you end up 
getting too few German families— and never know that for your purposes 
the sample is distorted. With a good probability sample, in contrast, there 
would have been a reasonable chance that Germans would have been drawn 
in proportion to their incidence in the population. 

(4) It is impossible to figure confidence limits on a quota sample, be- 
cause confidence limits are based on the free operation of chance. 
On the other hand, some of these deficiencies can be overcome, or else are 
not serious in relation to the marketing purpose for which the test is being made. 
And quota samples of large size are usually less expensive than probability samples 
of the same size. 

Thus, even though confidence limits cannot be computed, it is possible to do 
a rough job of deciding whether a sample is large enough to wash out most of the 
erratic variation in the sample results. If one sample of 500 yields a finding of 75.4, 
and another sample of 500 yields a finding of 75.2, it looks as if 500 is a large 
enough sample. Or, a more formal procedure is to average batches of returns as 
they come in by days, or weeks, or 100's (or they can be batched randomly after- 
ward), and then see how many it takes before the curve of the cumulative aver- 
age smooths out. Thus, on a question such as "How many people use a toothbrush 
regularly?" the result might be (in batches of 100 returns) : 









Cumulative 




Batch 


Frequency 


Cumulative 


Number of 


Cumulative 


dumber 


in Batch 


Frequency 


Returns 


Average 


1 


75% 


75% 


100 


75% 


2 


89 


164 


200 


82 


3 


58 


222 


300 


74 


4 


82 


304 


400 


76 


5 


79 


383 


500 


77 


6 


63 


446 


600 


74 


7 


73 


519 


700 


74 


8 


81 


600 


800 


75 


9 


65 


665 


900 


74 



10 83 748 1,000 75 

It looks as if the true figure is somewhere around 74 per cent or 75 per cent, 
and that a sample of 600 is large enough. Remember, however, that this is no- 
where so rigorous as using the confidence level formula in a probability sample. 

ADVANTAGES. Yet, if we know what factors are significant for the market- 
ing decision that is going to be based on the test, then a quota sample makes a 
lot of sense— why not get the economy of greater numbers at less waste effort 
which will result from a purposeful focusing on those factors? In a very real sense, 
this is almost like stratifying. And if, then, an arbitrary design is also brought in 



PRODUCT POTENTIAL 161 

to determine what units are drawn in satisfying quotas, one out of the two major 
advantages of probability sampling is obtained, that is, the avoidance of subjective 
judgment by the sampler when it comes to the last-minute selection of units to be 
drawn. 

In fact, there is much to be said for what is called "modified probability" 
sampling. Here, for example, the quotas of age, geographical location, income 
class, and so forth, are carefully fixed for a large area sample. But the on-the-spot 
interviewers have to follow rigid rules for the houses they are to call at, exactly 
as prescribed in probability sampling. In getting on the "probability" bandwagon, 
many researchers overlook some of the more practical but less "ideal" ways of 
getting a job done— which nonetheless may still be scientific, depending on how 
they are applied. 

Here, as everywhere else in marketing, adjusting the means to the nature and 
magnitude of the task at hand is just as scientific as always insisting on the 
ideal method— and, indeed, this adjustment often takes a great deal of analytical 
reasoning. 

Particularly on matters like the potential market for a new product, and the 
effect of quality, price, kind of outlet, promotion, and so forth, on its success, re- 
search can be useful— highly useful. But it is perforce imperfect and inconclusive, 
in large part because it is impossible to know in advance all the existing influential 
factors, let alone the new ones that are still to emerge; and even if it were possible 
to have this omniscience, it would cost too much, or take too much time, to 
measure them. This is another reason why a series of modest tests , repeated as the 
product and the market develop, makes so much more sense than a grand project, 
with all the blind confidence it engenders just because of its grandness. 

This is particularly true when, as we shall see in the next section, manage- 
ment engages in purposeful action to expand people's needs and wants— to create 
something new that can't be exactly measured today. And, as mentioned earlier, 
no matter how good the research, it is no substitute for the creative decision 
making and risk taking that management must do. 

Yet the techniques just considered may be of more relative value competitively 
than is commensurate with their relative contribution to a given decision or pro- 
gram. The reason is simply that the manager who makes better use of the informa- 
tion provided, may pull ahead of his equally creative competitor who is not as 
well provided with useful data to serve him as clues, as directions, as dimensions, 
and as correctives or confirmers of action in process. 



RESEARCH PLUS 

The foregoing discussion does not mean, however, that the problem of se- 
lecting new products and deciding whether to go ahead with them can be settled 
by marketing research, rough or refined. Far from it. 



162 PRODUCT POTENTIAL 

First of all, there is still need of logical analysis of the pull-apart-the-problem-by- 
questioning kind. This is indispensable, just to know what can be and what should 
be further subjected to marketing research. 

There are a number of ways in which such analysis applies: 

(1) Initial common-sense factoring of the particular product proposed: 
'Will our existing reputation or brand help it and/or be helped by it?" "Is this 
something our salesmen (including brokers, wholesalers, retailers, and so forth) 
can sell and will be interested in selling?" "Are the kind of outlets and chan- 
nels of distribution we now use suitable and helpful for this product?" "Do 
such things as credit terms, availability of advertising media, availability of 
servicing organization, spare parts stocking, and so forth, raise any difficul- 
ties?" The last is a particularly important question when going into a new 
country. 

(2) Or, even more broadly, "What kind of product or products should 
we be considering in the first place, again in the light of such questions as just 
posed under 1?" 

(3) Or, as implied by 2, "What kind of over-all strategy of product 
introduction, backed up by what kind of organization, should we have?" 
"Should we hold new offerings to three or four major products per year, on the 
ground that more than that many heavy promotional efforts might dilute our 
advertising impact and our field selling effectiveness?" "Do we want a broad 
line, so as to increase the number of prospective customers and/or the extent 
of needs that we can take care of for present customers; or a specialty line 
that we can push in depth because of its new or advantageous or unique 
features?" "Do we really want to get into the thick of the competitive fray, 
and compete on price and wheeling-and-dealing; or do we want to make a 
higher rate of profit on less volume, by appealing to particular users?" and 
so on. (These are top policy problems, and as such are involved in the mat- 
ters discussed in Section Six.) 

Remember that when a company increases business by adding new products, 
it is inevitably getting into new fields where, initially at least, it may be less pro- 
ficient than its competitors, or may be able to offer only "me-too" products. Hence 
it is necessary to consider the major alternative of putting the same amount of 
effort into expanding sales of present products. The new product area often looks 
more glamorous simply because less is known about the difficulties. But that 
doesn't mean it should not be considered. The pay-off is greater, many times. Only, 
in being considered, it should be compared against the possible rewards from push- 
ing existing products, as discussed in Section Four. 

* * * 

In the cases which follow, the focus is strongly on the question of how 
management should appraise the research methods which produce the findings 
about product potential that it must translate into action, as it makes its decision 
about going ahead with the product. 



PRODUCT POTENTIAL 



163 



In October, 1956, the Johnson Market- 
ing Research Agency was asked to deter- 
mine whether a fresh cranberry-orange 
mixture in crush form could be profitably 
marketed by the Pure Test Farms Dairy. 
Executives of the dairy believed that any 
new product that could be distributed with 
conventional dairy items along established 
milk routes offered a means of additional 
income at little added expense. The fact 
that the cranberry mixture would keep 
fresh for two to three weeks under nor- 
mal refrigeration also made Pure Test 
executives feel that the new item would 
be compatible with dairy products. 

The cranberry-orange mixture was made 
from the following ingredients: cranber- 
ries, 63 per cent; sugar, 25 per cent; and 
whole oranges, 12 per cent. Similar re- 
cipes were known to be prepared at home 
by housewives. The dairy marketing de- 
partment planned to distribute the item 
in reusable plastic containers containing 
10.5 oz. 

The product could be served with meats 
or poultry, as a salad and for dessert. 
Dairy executives believed its applications 
were restricted only by the imagination of 
the user. 

Costs of the market research project 
were borne jointly by Pure Test Farms, 
which distributed product samples, and 
by the National Cranberry Association, 
which provided the samples and under- 
wrote other costs of the research. The 
survey was conducted in Worcester, Mas- 
sachusetts, and in 13 towns between 
Worcester and Boston. 

The researchers believed that, in ad- 
dition to measuring product acceptance, 



CASE 3-1 

Pure Test Farms Dairy (A) 

their study should get answers to the fol- 
lowing questions: 

1. What price will people pay for this 
product? 

2. What is the most acceptable name? 

3. What promotional appeals will best 
stimulate sales? 

4. How could the product be improved? 
The researchers decided to conduct the 

survey by means of a mailed question- 
naire. They believed this would be the 
most satisfactory method not only from a 
cost standpoint but also from the stand- 
point of effectiveness. They felt the infor- 
mation could be best obtained right after 
the respondents had consumed the prod- 
uct, and a normal family meal would pro- 
vide a better environment than would be 
possible with an interviewer present. 

The questionnaire (see Figure 3-A) was 
finalized after slightly more than three 
weeks of work. After some debate the re- 
searchers decided to omit a question bear- 
ing on income. While they believed in- 
come data would be desirable because they 
felt cranberries were a high-cost ingredient 
and the product would be a luxury item 
for many families, they were also aware of 
the opinion, held by some experts, that 
questionnaire returns suffer when an in- 
come question is included. Also, Pure Test 
executives rejected the idea of asking cus- 
tomers to reveal annual income. Thus, in- 
stead of including an income question, the 
researchers planned to tour the sample 
area to obtain an approximate measure of 
economic status. However, the size of the 
area involved made this impractical, due 
to cost and time limitations. 



164 PRODUCT POTENTIAL 



PURE TEST FARMS DAIRY (A) 

Cranberry -Orange Product 

Questionnaire 

To assist us in creating cranberry products to your taste, we are presenting a new combination 
for your consideration. In order to obtain your opinion and the opinions of your family, we ask 
your cooperation in answering a few questions. With your family present, serve the product and 
THEN complete the questionnaire. If you complete the questionnaire and mail it within three 
days, we will say "thank you" by sending you a FREE Cranberry Sauce cutter. 

1. How do you like the taste of this Cranberry-Orange product? 

Like it Do not Strongly 

very much Like it Indifferent like it dislike it 

a. Wife: 

b. Husband: . 

c. Children: 

d. Other Household Members: 



2. Why do you like (dislike) this product? 
a. Wife: (age ) 



b. Husband: (age ) 




c. Children: (ages , , ) 




d. Other Household Members: (ages , 


) 


3. How could this product be improved? 





Based on family reaction and a reasonable price, would you buy this Cranberry-Orange 
product? 



Yes No 



5. If yes, about what price would you pay for this product? £ 

6. How did you serve this Cranberry-Orange product? 

With Meat Combined with Cottage Cheese 

With Poultry As a dessert 

As a salad As a snack 

Other (specify) 



What in your opinion is the most descriptive name for this product? 
"Cranberry-Orange 



(such as crush, mix, relish, compote or your suggestion) 

If you would care to receive a free Cranberry Sauce Cutter, please give your name and address: 

Name 



Address 



City_ State 

Figure 3-A 



PRODUCT POTENTIAL 



165 



In designing the questionnaire, the re- 
searchers attempted to make sure that 
questions were explicitly phrased and not 
overly time-consuming. In view of the 
importance of family influence on food 
buying decisions, a reaction from each 
household member was desired. It was 
anticipated that the housewife would 
take charge of filling in the blanks and 
that she would treat the test as a game. 
To promote her cooperation, the NCA 
provided free cranberry sauce cutters as 
a reward for prompt return of the ques- 
tionnaire. Stamped envelopes were in- 
cluded, bearing the address of the NCA. 

The product samples and the question- 
naires were divided among 15 milk-route 



drivers who delivered them to home-de- 
livery customers whom the drivers chose 
at their discretion. 

The universe sample size was set at 
1,000 families. 



Is this test likely to give National Cran- 
berry Association the information it needs 
to decide whether to market the new 
product and, if so, what to do about 
price, name, promotion, and product im- 
provement? 

Will it be of help to Pure Test Farms 
Dairy? 



166 



PRODUCT POTENTIAL 



CASE 3-2 

Consumer Attitudes Toward 
Irradiated Food Products 



The primary objective of this project 
was to find out what the attitude of food 
buyers (mainly housewives) was toward 
irradiated food products and how they 
would react when these became available 
in the market. According to the researcher: 

If sufficient time and money had been 
available, the ideal sample design would 
have been a random, or probability, sam- 
ple of all food buyers in the United States. 
However, since neither the time nor the 
money was available, the sampling unit 
chosen had to be considerably smaller. 
Therefore the city of Cambridge, Massa- 
chusetts, was chosen and the problem 
became one of determining the attitudes 
of the food buyers in the city of Cam- 
bridge toward irradiated food products. 

A random sample was chosen over a 
nonprobability sample for three reasons. 
First, a random sample would be less 
biased; second, the relative efficiency of 
sample design could be judged; and, 
finally, detailed information about the 
composition of the universe would not be 
required. 

The biggest problem presented, how- 
ever, was the need for some type of list 
which would identify all items in the 
universe. Such a list was not available. 
The possibility of using the phone book 
was quickly discarded, since it was clear 
that all families without phones would be 
excluded from such a list. A list of regis- 
tered voters would also be incomplete, 
and even the block list of the U. S. 
Census would not provide the names and 
addresses of all the food buyers in Cam- 
bridge. 



Therefore, the researcher decided to 
evolve a plan of his own, which would 
be simple, inexpensive, and provide "rea- 
sonable randomness." The first step was 
to obtain a map of Cambridge; over this 
map was placed a grid with numbered 
coordinates running along each axis (see 
Figure 3-B). 

A published table of random numbers 
was then utilized to specify locations on 
the map. For example, when the random 
number 90087 came up, it was interpreted 
to mean the point at which line 90 on 
the horizontal axis intersected line 087 
on the vertical axis. A pin was then 
pushed through the grid onto the map 
at this point, which turned out to be the 
northeast corner of Brookline and Putnam 
Streets. When 50 points had been lo- 
cated, they were numbered from one to 
50. The point just described was num- 
bered point #45, and was to be the 45th 
house at which the interviewer would 
call. 

There were more than 50 points on the 
map, however, because a few of them 
were unusable. For example, the two in 
the upper left-hand corner of Figure 3-B 
fell into the center of Fresh Pond, where 
it was obvious that the interviewer would 
find no food buyers to interview. 

The next question that arose was what 
to do when the exact house was not as 
clearly indicated as it was at point #45. 
For example, point #26 (at coordinates 
53-085 in Figure 3-B) indicates a house 
in the center of the block on the north 
side of Cleveland Street, but it is not 
clear whether this might be the 3rd, 4th, 



RTTTTEra 







o 


O 


o 


o 


o 


o 


o 


o 


o 


CO 


to 


*• 


m 


tf> 


^ 


00 


0> 


o 



SIXV -IV0IJ.U3A 



168 



PRODUCT POTENTIAL 



or even 5th house in the block. (It would 
be clear if this were a larger scale map, 
however.) One solution might have been 
for the interviewer to make a choice when 
he reached the location. However, bias 
could become a serious factor here. If the 
interviewer were allowed to make value 
judgments, he would probably tend to 
choose the "easiest" house to interview, 
thus perhaps excluding houses with such 
impediments as dogs and high fences. He 
might also tend to pick better looking or 
cleaner houses, thus excluding the lower 
income groups. Therefore, it was decided 
that the interviewer would carry dice in 
his pocket. When he reached the loca- 
tion, he would roll them, and if number 
four came up, he would interview the 4th 
house from the corner. 

The same system was used in multiple 
dwelling units to select the apartment to 
interview. Here it was recognized by the 
researcher that he was departing from 
"perfect randomness," since, in theory, a 
person living in a multiple dwelling unit 



has less chance of being selected than a 
person living in a single dwelling unit. 
The "ideal solution" would have been to 
equalize the apartment dweller's chances 
by interviewing several people in the apart- 
ment in contrast to one in the individual 
home. 

However, to do this correctly would 
have entailed increasing the size of the 
sample considerably beyond the limits im- 
posed by time and funds. In spite of this 
fact, the researcher did not feel that the 
bias introduced here was significant. 



Is this a satisfactory sample design, 
from the standpoint of size, universe, 
listing, and method of selection? 

What flaws are there from the strict 
statistical point of view? Are they serious? 
Does the purpose of the test make any dif- 
ference in the answers to any of these 



questions? 



PRODUCT POTENTIAL 



169 



In August, 1956, Spalding Market Re- 
search Associates, a local marketing re- 
search firm of Boston, Massachusetts, 
received an assignment to investigate the 
TV viewing habits of the residents of 
nearby Cambridge. 

The client, a large TV network, wanted 
Spalding to undertake an exploratory study 
to determine what types of people pre- 
ferred what types of programs. The TV 
network executives believed that if "type" 
of person were interpreted as "personality 
type" rather than as type by age, sex, 
socio-economic, or employment classifica- 
tion, information on program or program 
content preferences was not available. 
These executives also believed that such 
information might be valuable both to 
the network and to its advertisers. If 
successful in Cambridge, the TV network 



CASE 3-3 

Spalding Market Research 

Census Bureau's "Block Statistics" for 
Cambridge, a sample page of which is 
reproduced in Figure 3-C. He grouped all 
the blocks in Cambridge into five cate- 
gories, according to the number of dwelling 
units in each, and compiled the table 
below. 

On the basis of experience, Mr. Spald- 
ing decided to obtain five interviews from 
each block that entered into his sample. 
Therefore, in order to have 150 interviews 
it would be necessary to have 30 blocks in 
the sample. Mr. Spalding wanted to have 
two additional blocks available in case of 
unforeseen contingencies, however, so he 
set the total number of blocks wanted at 
32. Part of the Census Bureau's map of 
Cambridge is reproduced in Figure 3-D. 

Mr. Spalding said that because he 
wanted each dwelling unit in Cambridge 





Number of dwelling 


Median of the 


number 


Number of blocks 




units per block 


of houses in group 


in group 


Group I 


- 25 


12.5 




208 


Group II 


26 - 50 


37.5 




213 


Group m 


51 - 100 


75.0 




192 


Group IV 


101 - 150 


125.0 




41 


Group V 


151 and above 


185.0 




22 



planned to conduct similar studies in 
other cities. 

Spalding was chosen because it was 
known to have the services, on a consult- 
ing basis, of a capable psychologist and 
because it specialized in pilot studies. 

Because of a small budget and the lim- 
ited availability of interviewers he consid- 
ered qualified, Mr. Spalding decided to 
keep the size of the study down to no 
more than 150 interviews. 

Mr. Spalding developed a sampling plan 
for the study. He first consulted the 



to have an equal chance of being part of 
the sample, he had to include a larger pro- 
portion of the blocks with a high density 
of dwelling units than of those blocks 
with few dwelling units. His method of 
achieving this weighting is shown by the 
following calculation: 



Group I 


12.5 


x 208 = 


2,600 


Group II 


37.5 x 213 = 


7,988 


Group III 


75.0 


x 192 = 


14,400 


Group IV 


125.0 


x 41 = 


5,125 


Group V 


185.0 


x 22 = 


4,070 


Total 






34,183 



CAMBRIDGE, MASS. 





Tabic 3.- 


-CHARACTERISTICS OF HOUSING FOR CENSUS TRACTS. BY BLOCKS: 


1950-Co 


n. 






Block 


All dwelling unite by occupancy 
and tenure 


All dwelling unit* 
by condition and 
plumbing facilities 


Occupied dwelling units 


Contract monthly 
rent 1 


Value' of one-dwelling- 
unit structures 


OlWUD 

tract 


Totel 


Owner 
occupied 


Renter 
occupied 


Vacant 
non- 
seasonal 

not 
dilap., 
for rent 
or sale 


Other 
vacant 
and 
non- 
resident 


Number 
reporting 


No 
private 
bath or 


No 
running 


Total 


Persona 
rooo 

Number 
reporting 


per 
> 

1.51 
or 
more 


Occupied 
by non- 
white 


Number 
reporting 


Average 
monthly 

rent 
(dollars) 


Number 
reporting 


Average 

value 
(dollars) 




water 

or 
dilap. 


MC-12 


37 


41 


12 


29 






39 


1 


1 


4 1 


40 


1 




29 


3840 


5 


10.1 


MC-13 


1 


5 5 


14 


4 1 






55 


2 


1 


55 


55 


1 




41 


4 5.7 


9 


855 5 




2 


44 


6 


38 






4 4 


22 


2 


4 4 


4 4 


4 




38 


3628 








3 


33 




32 


1 




33 


4 


3 


32 


33 


1 




33 


2 8.8 7 








4 


60 


8 


47 




5 


59 


6 


2 


55 


54 


2 


3 


45 


4 6.80 


4 


6400 




5 


42 


14 


28 






4 2 


5 


1 


4 2 


43 




1 


2 5 


4 5.4 4 


6 


758 3 




6 


69 


20 


49 






67 


ai 




69 


69 


2 


4 


4 9 


3 6.81 


S 


51 




7 


96 


18 


76 




2 


94 


14 


3 


94 


93 


1 


13 


76 


4 6.7 1 


8 


5.12 5 




8 


101 


16 


85 






101 




3 


101 


101 




1 5 


83 


3958 


4 


423 5 




9 


41 


9 


32 






39 






41 


39 


3 




31 


3 9.61 


2 






10 


17 


4 


1 3 






1 7 






17 


17 






1 3 


2 9.7 6 


1 






11 


52 


11 


4 1 






51 




4 


58 


52 






4 1 


4 05 5 


2 






ia 


124 


20 


104 






12 4 


13 


5 


124 


122 


3 


1 2 


104 


3 9.7 7 


10 


635 




13 


3 5 


10 


25 






34 






35 


34 


1 




24 


4 012 


2 






14 


56 


12 


4 4 






54 






56 


53 


1 


4 


40 


4 255 


3 


646 6 




1 5 


68 


11 


57 






68 


18 




68 


68 


3 


7 


56 


3 216 


1 






16 


83 


10 


71 


1 




81 






81 


81 




3 


72 


3 0.4 7 


4 


327 5 




17 


51 


18 


3 3 






51 




3 


51 


Si 






3 2 


4 1 12 


2 






18 


68 


13 


55 






68 






68 


68 


1 




53 


5 0.13 


2 






19 


33 


7 


1 5 






20 


5 


2 


22 


22 




1 


1 5 


4753 


1 






ao 


9 


2 


7 






8 






9 


9 






7 


2814 








ai 


43 


8 


34 






4 2 


a 


8 


42 


42 




2 


3 3 


2 7.42 


3 


64 




32 


30 


12 


1 7 




1 


28 






29 


28 






16 


3 1.31 


1 






23 


60 


IS 


4 5 






60 


2 




60 


59 






41 


4 6.4 1 








34 


38 


7 


31 






38 






38 


38 






31 


4 1.61 


1 






35 


48 


IS 


36 


1 




48 


5 


4 


47 


47 






36 


3 3.05 


1 






26 


90 


78 


2 




89 


3 


3 


88 


87 


3 




79 


38.50 


1 






37 


76 


19 


57 






75 






76 


76 




2 


57 


3 5-22 








38 


9 


4 


5 






9 






9 


9 






5 


4 9.00 


1 






39 


32 


11 


1 


1 




22 


3 


3 


21 


21 






1 1 


3718 


5 


730 




30 


44 


9 


34 


1 




4 4 






43 


4 3 






34 


3440 


1 






31 


47 


16 


30 




1 


47 


1 


1 


46 


46 






28 


3714 


3 


7.3 3 3 


MC-1 4 


1 


1 


































a 


1 36 


25 


111 






133 


17 


7 


136 


132 


5 


5 


107 


3 5.26 


3 


1 0.00 




3 


14 


2 


1 2 






1 4 






1 4 


1 4 


1 




12 


3048 








4 


9 


1 


8 






9 






9 


9 






8 


3140 








5 


39 


10 


29 






39 


1 


1 


39 


39 




5 


28 


2817 


3 


8400 




6 


23 


2 


20 






22 






22 


22 


1 


1 


20 


3 8.7 5 








7 


83 


8 


15 






a 3 






23 


23 


a 


2 


1 4 


2647 


4 


332 5 




8 


44 


15 


38 




1 


43 






43 


4 3 




4 


27 


2940 


1 






9 


43 


9 


34 






4 3 


2 


2 


43 


4 3 


l 


7 


3 4 


2611 


5 


4200 




10 


50 


6 


4 3 




1 


49 


20 


11 


49 


49 


5 


10 


4 3 


2 4.60 








11 


48 


13 


35 






48 


7 


3 


48 


47 


3 


20 


3 5 


2 3.8 5 


6 


633 3 




13 


38 


9 


38 


1 




38 


7 


5 


37 


36 


1 




29 


2 4.7 9 


1 






1 3 


18 5 


33 


152 






185 


46 


35 


185 


185 


7 


6 1 


152 


2 531 


5 


7.9 




14 


20 


5 


15 






30 






20 


20 




3 


1 4 


2647 


2 






15 


6a 


7 


53 


2 


1 


62 






59 


59 


3 


1 


53 


2 318 


2 






17 


7 




7 






7 






7 


7 


1 




7 


3140 






IIC-15 


a 


13 


3 


10 






1 2 


7 


7 


13 


12 






8 


46i0 


1 






3 


103 


14 


87 




1 


99 


68 


61 


101 


98 


7 


2 


85 


2252 


5 


4300 




4 


48 


11 


37 






4 8 


5 


1 


48 


48 


3 


10 


36 


3654 


7 


5.3 1 4 




5 


91 


33 


69 






90 


24 


19 


91 


90 


2 


24 


69 


22.69 


2 






6 


47 


15 


3 1 




1 


45 


3 


2 


46 


4 5 


3 


7 


29 


26.65 


4 


7.7 5 




7 


89 


24 


65 






89 


13 


7 


89 


89 


4 


6 1 


63 


2 4.68 


4 


4400 




8 


31 


11 


1 8 




2 


28 


3 


3 


29 


29 


1 


1 6 


1 8 


2 3.7 2 


4 


5.3 7 5 




9 


39 


5 


3 3 




1 


39 


32 


31 


38 


38 


2 


5 


3 3 


2 IIS 


1 






10 


33 




19 




3 


30 


16 


16 


19 


1 9 


1 




1 9 


2 4.7 3 








11 


101 


22 


79 






101 


30 




101 


101 


3 




79 


2645 


9 


7.3 3 3 




13 


78 


23 


50 


3 


3 


78 


13 


2 


73 


7 3 


1 


9 


4 9 


2 4.26 


7 


54 4 2 




13 


8 


18 


60 


1 


1 


7 9 


1 2 




78 


7 7 


1 


1 8 


58 


2 5.4 8 


5 


450 




14 


59 


9 


48 




2 


59 


3 




57 


57 


4 


3 8 


4 8 


2 2.54 


2 






15 


38 


11 


17 






37 


2 




26 


27 


3 


1 5 


16 


25i0 


4 


4.1 2 5 




16 


51 


6 


45 






51 


9 




51 


5 1 


1 


4 5 


4 5 


2 422 


1 






17 


69 


5 


64 






69 


8 


8 


69 


69 


3 


3 8 


64 


2 3.34 








18 


19 


5 


14 






1 9 






19 


19 




1 4 


1 4 


2 1.7 1 


1 






19 


39 


9 


30 






39 


20 


20 


39 


39 




1 9 


30 


2410 


1 




MC- 16 


1 


37 


1 


36 






27 






27 


27 






26 


53.30 








a 


19 




1 8 




1 


1 9 






18 


18 






18 


9841 








3 


3 


3 








3 






3 


3 










3 


2 3 .3 3 3 




4 


1 5 


5 


1 






1 5 


1 




1 5 


1 5 


1 




10 


8 1.7 


3 


2 2.3 3 3 




5 


39 


13 


16 






2 9 






29 


29 


1 




1 6 


7 946 


5 


lfi^OO 




6 


ao 


11 


9 






20 






20 i 20 








1 1 7.4 2 


10 


2 5 5 




7 


10 


6 


4 






10 






10 


10 








9 OjOO 


2 






8 


3 


2 


1 






2 






3 


3 










1 






9 


111 


23 


87 




1 


111 


1 


1 


110 


110 


2 




8 1 


60.38 


1 8 


2236 6 




10 


33 


9 


2 3 




1 


33 


1 




32 


32 


2 




2 3 


3 5.13 


1 





'For renter - occupied dwelling units and vacant nonseasonal not dilapidated unite, for rent. 
'Fox owner - occupied dwelling unit* and vacant nonseasonal not dilapidated units, for sale only. 



Figure 3-C 



PRODUCT POTENTIAL 



171 



CAMBRIDGE. MASSACHUSETTS. 
BY CENSUS TRACTS AND BLOCKS: 1950 



PART 2 OF 3 PARTS 




B LOCK NUMBERS 

TRACT NUMBERS 

TRACT BOUNDARIES _— — __-— — . -. 

U.S. DEPARTMENT OF COMMERCE BUREAU OF THE CENSUS 

Figure 3-D. 



172 



PRODUCT POTENTIAL 



Since the product of the median of 
the group and the number of blocks 
in the group when totaled for all groups 
equaled 34,183, Mr. Spalding stated that 
the number of blocks to be included in 
the sample from each group could be 
determined by dividing 32 into 34,183, 
obtaining approximately 1,060, and then 
dividing the 1,060 into the product for 
each group as developed in the above 
calculations. 

On this basis, Mr. Spalding determined 
that the number of blocks from each 
group to be included in the final sample 
should be as follows: 



Group I 


3 


Group II 


7 


Group HI 


13 


Group IV 


5 


Group V 


4 



a table of random numbers to select three 
by chance. He performed a similar opera- 
tion for the other groups. 

Mr. Spalding planned to have 10 inter- 
viewers. Three blocks were to be assigned 
to each interviewer and the interviewers 
were to select at their discretion any five 
dwelling units within that block. Mr. 
Spalding decided that only those 18 years 
of age or older were to be interviewed 
and that students of any age were to be 
excluded from the sample. Beyond these 
stipulations, interviewers were to be free 
to select respondents within the dwelling 
units they selected, provided that each 
interviewer had completed interviews with 
approximately the same number of men 
and women respondents when his or her 
assignment was finished. 



To determine just which three blocks 
should be included from Group I, for ex- 
ample, Mr. Spalding numbered all the 
blocks falling into that category and used 



Is this sample design satisfactory? What 
criteria should be used in determining its 
usefulness? 



PRODUCT POTENTIAL 



173 



Two trainees in a large marketing re- 
search agency made the following study 
and report for the benefit of their own 
firm: 



Objective 

To compare, and to attempt to explain, 
the differences arising from the responses 
to an identical questionnaire administered 
in three different ways, to groups chosen 
by the same probability sampling method 
from the same universe. 



Discussion of the Objective 

It is surprising to discover the scant 
information, in the field of highly per- 
sonal questions, available to a practical 
businessman should he be desirous of 
conducting interviews in which these 
questions arise. While it is true that psy- 
chological methods, such as projective 
techniques, depth interviews, and so on, 
have helped to answer some of these 
questions, to many businessmen they still 
have the following limitations and disad- 
vantages : 

1. They require an expert in psychology 
to construct and interpret them, with 
the result that most businessmen have 
little idea of just exactly what is hap- 
pening in the tests, and are conse- 
quently uncertain of the validity and 
interpretation of the results. 

2. The cost to conduct a survey is rela- 
tively expensive. 

3. It is impossible to get the definite re- 
sults obtained through the use of a 
questionnaire. 



CASE 3-4 

La Crosse Marketing Research Agency 

Hence, the questionnaire still represents 
the most commonly used method of col- 
lecting data, and is the one method most 
easily understood by both businessmen 
and researchers. However, many of these 
questionnaires contain highly personal 
questions, similar to questions No. 6 to 
No. 11, inclusive, on Figure 3-E, and 
oftentimes management wonders whether 
or not people will be truthful in answer- 
ing those questions and whether or not 
a different surveying technique would 
have produced different responses to those 
questions. The aim of this project was to 
investigate this latter point and can be 
stated as follows: 

1. Do women answer certain types of 
highly personal questions differently, 
depending upon the surveying tech- 
nique used to obtain the responses? 

2. If differences do arise, which method 
of interviewing investigated gives the 
most valid results? 

Because a questionnaire was being used, 
the types of surveying methods to be in- 
vestigated followed almost automatically, 
and are: personal interviewing, telephone 
interviewing, and direct mail. The ques- 
tions were, of course, identical for all 
surveys, thereby reducing the number of 
variables introduced into investigation. 

Why should a woman answer questions 
differently depending upon the surveying 
technique used to obtain these answers? 
It is thought that the type and the con- 
tent of the question have a great deal to 
do with this. In certain questions, such as 
marital status of interviewee, and the like, 
it is very, very doubtful that the inter- 
viewee would not tell the truth, because 



174 



PRODUCT POTENTIAL 



she has no motive to respond differently. 
However, in more personal questions, such 
as "Do you wash your dishes after every 
meal?" it is highly probable that a woman 
might tell a "white lie" in order to give 
a more favorable impression of herself. 
Therefore, under which surveying method 
would a woman be inclined to give a more 
favorable impression of herself? 



Sampling 

Location. Time would not allow us to 
sample adequately a large area. Hence, the 
area was deliberately restricted to a size 
which, in the time allotted, would allow 
the interviewers to obtain representative 
subsamples of the universe. Evanston, Illi- 
nois, was chosen as this area because of its 
proximity to the research firm's office, thus 
minimizing both the cost and the time of 
phoning and transportation, and because 
it allowed us to pose as students. 

The Universe. From the universe, con- 
sisting of women having a resident tele- 
phone listing in Evanston under either 
their own or their husband's name, three 
samples were drawn at random — 108 for 
the direct mail survey, 108 for the tele- 
phone interview survey, and 27 for the 
personal interview survey, a total of 243. 
On every fourth page of the Chicago sub- 
urban directory, the researchers started at 
the extreme right hand column and took 
the name, address, and telephone number 
of the first listing with an Evanston ex- 
change prefix which was not a known 
student dormitory, boarding house, or busi- 
ness number. If no number was found in 
this column, the authors started at the 
foot of the next column and worked back- 
ward through the column (s) until a suit- 
able number was obtained. This procedure 
was repeated until 243 names were ob- 
tained. 

The names, listed in alphabetical order, 



were numbered 1 to 243 and were then 
split into groups of 1 to 9, 10 to 18, and 
so forth, yielding 27 groups of 9 names. 
Each group of 9 was further divided into 
groups of 4:4:1. The first four names 
were used in the direct mail survey; the 
second four names were used in the tele- 
phone survey; the last name was used in 
the personal interview survey. 

Some of the details of the actual ques- 
tioning were: 

Personal Interviewing. Of the original 
27 people chosen for the subsample of 
personal interviews, 20 women supplied 
the information within the two allowed 
call-backs; three women refused to coop- 
erate; two women were still out after two 
call-backs; one woman had moved to Cali- 
fornia; and one woman had died. To an- 
swer Question No. 16 and No. 17, the 
interviewee was handed a card, shown in 
Figure 3-F, and was asked to indicate by 
letter the range in which her income and 
age lay. 

Telephone Interviewing. In only about 
20 per cent of the cases were the inter- 
viewees at home on the first call, and, in 
some cases, as many as four call-backs 
were made. This method of interviewing 
posed some rather unique responses. For 
instance, reasons for refusing to partici- 
pate in the survey were: "My wife no 
speaka English"; "We are leaving for 
abroad in two weeks and don't want any- 
thing"; "When people ask these kinds of 
questions they come to the door"; "I'm 
sorry, I'm watching TV" (this happened 
three times with one hoped-for inter- 
viewee); and, "If you want to ask me 
questions over the phone, you will have to 
send me a letter first." Because of situa- 
tions such as these, only 79 of the initial 
108 interviews were obtained. In general, 
however, women did not object to being 
called on the phone and asked a long 
string of personal questions. Originally, it 



PRODUCT POTENTIAL 



175 



had been expected that there would be as 
many telephone tumdowns as unreturned 
questionnaires. 

Direct Mail Survey. The questionnaire 
is shown in Figure 3-E. Within two weeks, 
38 questionnaires were returned out of the 
original 108. 



Questionnaire Design 

The first step in constructing the ques- 
tionnaire was to choose questions of a 
type which women would tend to answer 
differently depending upon the interview- 
ing technique used. The type of questions 
which would amplify those differences 
would, it was thought, be related to un- 
desirable habits, socio-economic status, in- 
telligence, income, and age. Income and 
age were treated separately, and a list of 
questions relating to the other categories 
was compiled. These questions were then 
graded by asking five women to read each 
question, and to rank it according to the 
extent to which she thought other women 
would tend to give answers depending 
upon the interviewing technique used to 



obtain that answer. The questions with 
the highest rankings, that is, those which 
would amplify most the differences in 
responses due to the differences in inter- 
viewing techniques, were used in the ques- 
tionnaire. These are Questions No. 6 to 
No. 11, inclusive, on the questionnaire in 
Figure 3-E. It can be seen that these ques- 
tions are dichotomous; this was because 
the researchers wanted to force a definite 
answer to each of these questions. To in- 
clude a "Don't Know" category would, it 
was thought, lead to hedging on the part 
of the interviewee. 

Questions 1-5 were included simply to 
establish the equivalency of the three sub- 
samples. Questions 12-17 were included for 
classification purposes, so that further 
analysis could be made if desirable. 

Results are shown in Table 3-G. 



What do the findings indicate? How 
can the differences in results of the three 
methods be explained? Do questions No. 
16 and No. 17 have any bearing on the 
subject of the test? 



THE QUESTIONNAIRE 



1. Are you married? 

2. If married, how long have you been married? 

3. If married, how many children do you have? 

4. How long have you lived in the vicinity of Boston? 

5. On the average, how many meals do you cook each day? 

6. Do you wash your dishes immediately after every meal you cook? 

7. Do you serve instant coffee to guests? 

8. Do you ever read comic books? 

9. Do you read newspapers four or five times a week? 

10. Do you spend more money on clothes than you think you should? 

11. Do you particularly enjoy movies with love scenes? 

12. Do you or your husband own an automobile? 

13. How many years did you spend at school and college combined? 

14. How many rooms do you have in your house or apartment? 

15. How many people share these rooms? 

16. What is your approximate age? 

17. What was your approximate family income last year? 



Yes 


No 




yrs. 


children 


yrs. 


meals 


Yes 


No 


Yes 


No 


Yes 


No 


Yes 


No 


Yes 


No 


Yes 


No 


Yes 


No 




yrs. 


rooms 


people 


yrs. 


dollars 



Figure 3-E 



176 



PRODUCT POTENTIAL 



16. Would you mind circling your approximate 
age? 

A 18-21 D 31-35 G 46-50 K 61-65 N 76-80 
B 22-25 E 36-40 H 51-55 L 66-70 P 81-85 
C 26-30 F 41-45 J 56-60 M71-75 Q 86-90 



17. 



What was your approximate family in- 
come last year? (circle the approximate 
range) 

Income 



Yearly 




Monthly 




A $0,000 to $0,999 


or 


$000 to $83 


A 


B $1,000 to $1,999 


or 


$84 to $170 


B 


C $2,000 to $2,999 


or 


$171 to $250 


C 


D $3,000 to $3,999 


or 


$251 to $340 


D 


E $4,000 to $4,999 


or 


$341 to $415 


E 


F $5,000 to $5,999 


or 


$416 to $500 


F 


G $6,000 to $6,999 


or 


$501 to $584 


G 


H $7,000 and above 


or 


$585 and above H 



Figure 3-F. Cards Used in Personal Interviewing to Obtain Answers to Questions 16 and 17 



TABLE 3-G. INTERVIEWING TECHNIQUES 
Comparison of the average results of the three sub-samples 



ition mar, 


iber Mail 


Interview 


Phone 




n = 38 


n = 20 


n = 79 


(1) 


76% 


85% 


81% 


(2) 


16.2 yr. 


19.6 yr. 


18.0 yr. 


(3) 


2.5 children 


2.5 children 


2.1 children 


(4) 


32.5 yr. 


36.5 yr. 


27.0 yr. 


(5) 


2.4 meals 


3.05 meals 


2.5 meals 


(6) 


64% 


90% 


59% 


(7) 


39% 


40% 


52% 


(8) 


26% 


30% 


47% 


(9) 


92% 


85% 


81% 


(10) 


16% 


25% 


25% 


(ID 


39% 


21% 


52% 


(12) 


53% 


67% 


66% 


(13) 


13.8 yr. 


11.7 yr. 


10.4 yr. 


(14) 


5.4 rooms 


5.2 rooms 


6.3 rooms 


(15) 


3.3 people 


3.9 people 


3.8 people 


(16) 


44.2 yr. 


43.5 yr. 


45.0 yr. 


(17) 


$4,200 


$3,700 


$4,050 



PRODUCT POTENTIAL 



177 



CASE 3-5 

Pure Test Farms Dairy (B) 



Of 1,000 questionnaires distributed in 
connection with the market research pro- 
ject on the orange-cranberry product (Case 
3-1), 181 were returned by the cut-off 
date, December 18, 1956. 

Survey results appear in Table 3-H. 
* * * 

What guides to action, if any, can the 



National Cranberry Association obtain 
from the findings, in terms of (a) the 
acceptability of the product; (b) specific 
problems of price, name, promotional 
appeals, and use; (c) further research 
needed. Can the findings be projected to 
the national market? Do they need to be 
for Pure Test Farms Dairy? 



TABLE 3-H. PURE TEST FARMS DAIRY (B) 
Tabulation of Survey Results 

Question 1: How do you like the taste of this product? 

Like it Do Not 

Very Much Like it Indifferent Like It 

No. % No. % No. % No. % 

A. Wife 106 58.5 46 25.5 8 4.5 18 10 

181 answers 

B. Husband 74 46 47 29 19 12 15 9.5 

161 answers 

C. Children 62 24.5 80 32 29 11.5 71 28 

252 answers 

D. Other 19 63 4 13.5 4 13.5 

Total 261 42 177 28.5 56 9 108 17 

624 answers 



Strongly 
Dislike It 
No. % 



3 

6 

10 

3 
22 



1.5 
3.5 

4 

10 
3.5 



Question 2: Why do you like (dislike) this product? 
Likes 



Children 



No. 



Tastes good 7 

Tangy taste 6 

Like cranberries in any form . . 6 

Like color 3 

Like flavor 3 

Good for sandwich 1 

Like combination 1 

Like with meals 1 

Like orange flavor 1 



24 

21 

21 

10.4 

10.4 
3.3 
3.3 
3.3 
3.3 



Dislikes 



No. 



Too sour 10 48 

Not consistent 4 19 

Don't like orange taste 2 9.4 

Don't like taste 1 4.75 

Too chewy 2 9.4 

Too lumpy 1 4.75 

Dislike rind _1 4.75 

Total 21 100.00 



Total 29 100.00 



TABLE 3-H (Continued) 



Adults 



Like flavor 15 16.5 Too much rind 9 22 

Different 15 16.5 Too tart 8 19.5 

Tasty 13 14.5 Too much orange (not enough 

Like Tanginess 12 13 cranberries) 7 17.1 

Like orange flavor: takes Don't like 5 12.3 

away cranberry tartness. ... 11 12 Too bulky 4 9.8 



16.5 


16.5 


14.5 


13 


12 


11 


5.5 


5.5 


4.5 


1 



Good combination 10 11 Can't eat oranges 3 7.3 

Delicious 5 5.5 Don't like cranberries 2 4.8 

It's a change 5 5.5 No different from other 

Like Mother makes 4 4.5 products 2 4.8 

Pleasant 1 1 Too sweet 1 2.4 



Total 91 100.00 Total 41 100.00 

Question 3: How could this product be improved? 

No. % 

Good as is 47 37 

Remove rind— more cranberry 32 25.5 

Sweeter 14 11 

Cut finer or make smoother 13 10.5 

Add nuts 5 4 

Prefer it jellied (same flavor) 4 3 

Add apple 3 2.5 

Less sweet 2 1.5 

Add apple but no rind 1 1 

Use orange juice in place of rind 1 1 

Add nuts and celery 1 1 

Add lemon juice 1 1 

More liquid consistency 1 1 



Total 125 100.00 

Question 4: Based on family reaction and a reasonable price, would you buy this product? 

No. % 

Answered "Yes" 139 82.5 

Answered "No" 29 17 

Answered "Maybe" 1 J3 

169 100.0 

Question 5: If "yes" to question 4, about what price would you pay for this product? 

Retail Price (10 1/2-oz. container) No. % 

25? 
35<? 
20<? 
29<? 
30<? 
39<? 
40<? 
22<? 
19? 
Reasonable 
45<? 
49<? 
50<? 
15<? 
21<? 
23<? 
31? 
33? 
Any price 



26 


29 


15 


12 


14 


11 


12 


10 


11 


9 


6 


5 


5 


4 


4 


3 


4 


3 


4 


3 


2 


1.5 


2 


1.5 


2 


1.5 


2 


1.5 


1 


1 


1 


1 


1 


1 


1 


1 


1 


1 


114 


100.0 



TABLE 3-H (Continued) 
Question 6: How did you serve this cranberry -orange product' 



No. % 

With meat 117 38.5% 

With poultry 74 24 

As a snack 24 8 

As a salad 23 7.5 

With cottage cheese 17 5.5 

As a dessert 13 4 

As a relish 10 3.5 

With fish 3 1 

On toast 3 1 

With ice cream 3 1 

On crackers 2 0.75 

With whipped cream 1 0.35 

With lemon jello 1 0.35 

On English muffins 1 0.35 



No. % 



On French toast 

As canapes 

Mixed with Waldorf salad. . 

On lemon sherbet 

In pastry turnover 

With cream cheese 

In muffins 

On tapioca pudding 

Upside down cake 

As shortcake, with nuts and 

apple 

In Christmas cookies .... 
As delicacy 



0.35% 

0.35 

0.35 

0.35 

0.35 

0.35 

0.35 

0.35 

0.35 

0.35 
0.35 
0.35 



Total 304 100.00 



Question 7: What in your opinion, is the most descriptive name for this product? 



No. 

Cranberry-orange Relish 71 

Cranberry-orange Crush 33 

Cranberry-orange Delight 18 

7 

4 

3 

3 

2 

2 

1 

2 



Cranberry-orange Compote 

Cranberry-orange Mix 

Cranberry-orange Sauce 

Cranberry-orange Conserve 

Cranberry-orange Supreme 

Crushed Cranberry-orange Relish 

Cranberry-orange Salad 

Cranberry-orange Salad Mix 

Cranberry-orange Combination 

Cranberry-orange Conserve Delight 

Cranberry-orange Crisp 

Cranberry-orange Relish Treat Supreme 

Cranberry-orange Mix Crush 

Cranberry-orange Holiday Crush 

Cranberry-orange Lush Relish 

Cranberry-orange Jubilee 

Cranberry-orange Surprise 

"C range" 

Orange-Cranberry Mix 

Delicious Cranberry Crush 

Cranberry-Orange Blend 

Cranberry-Orange Crunch 

Cranberry-Orange Delicacy 

Oran-berry Crush 

Total 162 



44.5 
20.5 
11.4 
4.5 
2.5 
2 
2 

1.2 
1.2 



0.6 
1.2 



0.6 

0.6 

0.6 

101.20 






179 



180 



PRODUCT POTENTIAL 



CASE 3-6 

Gem Appliance Company 



In 1961, Mr. Daniel Goodman, Mana- 
ger of the Gem Appliance Company, 
located in Tel Aviv, was considering add- 
ing an electric dryer to the line of home 
appliances he manufactured. This would 
sell at around IL. 300. With a washer at 
around IL. 400 and a separate dryer at 
IL. 300, complete laundry equipment 
would thus come to about IL. 700. 

Gem was already manufacturing elec- 
tric washing machines; and a competing 
model on the market, which operated 
automatically, included an air-circulating 
feature that dried the laundry partly but 
not enough to preclude the need for fur- 
ther drying or ironing (priced at around 
IL. 540). But dryers would be a new 
product for Gem, as for most people in 
Israel. So Goodman needed to forecast 
the possible size of the electric dryer 
market. 

All the information he had available 
was a study of the U.S. market for dryers 
when it was just opening up in 1948, 
made by the Research Department of 
The Curtis Publishing Company (which is 
shown, abridged, in the appendix); and 
some more general economic figures for 
Israel, drawn from the Statistical Abstract 
of Israel, a government publication, and 
from the Statistical Yearbook of the 
United Nations (see Table 3-1). Mr. 
Goodman also learned from Electrical 
Merchandising Week that in 1960, in the 
United States, 3,274,000 electric washers 
were sold, and 1,238,000 electric and gas 
dryers. Further inquiry revealed that this 
ratio prevailed without much variation 
in all regions of the country 



Should Mr. Goodman go ahead with 
plans to add the dryer to his line of ap- 
pliances? What general factors will deter- 
mine the demand for this kind of ap- 
pliance? What factors that may be peculiar 
to Israel as against the United States? In 
particular, just what is the major buying 
motivation for the dryer in the United 
States, and will this prevail in Israel too? 
Are there important differences between 
washers and dryers, between dryers and 
ironers, and so on? 



Appendix — CURTIS RESEARCH DEPARTMENT 
BULLETIN 

Automatic clothes dryers are relative 
newcomers to the domestic appliance field 
but have received nationwide publicity 
since the end of the war. Experimental 
production of this device was begun as 
early as 1936, but relatively few dryers 
were sold to the public before war broke 
out and output was curtailed. Since the 
war, production has been resumed and the 
automatic clothes dryer is beginning to 
appear on the market in greater volume. 
Sales of 58,000 units having a retail value 
of $12,180,000 were reported in 1947, 
the only year for which figures are avail- 
able on an industrywide basis. Thus, it is 
well on the road to becoming a popular 
and valuable addition to home laundry 
equipment. But, to realize the ultimate 
national market potential, a consistent 
program of consumer education and adver- 
tising is necessary. . . . 

The basic operating principle of the 
dryer utilizes a revolving cylinder which 
tumbles the wet clothes about while they 
dry in warm air supplied by a heating unit 



PRODUCT POTENTIAL 



181 



TABLE 3-1. ECONOMIC STATISTICS FOR ISRAEL 



1949 



1951 



1953 



1955 



1957 



1958 



1959 



Population 
(In millions) 


1.2 


1.6 


1.7 


1.8 


2.0 


2.0 


2.1 


Education (In thousands 
of pupils) 
Total 


141 


249 


355 


381 


472 


517 


551 


Higher Education 


2 


4 


6 


7 


8 


10 


11 



246 



,308 



216 



259 

14,750 
15,490 



1,416 1,766 



265 

15,210 
17,880 

1,968 



Residential Building 

(Index) 100 293 138 

(Number of Dwellings) 
With 2 rooms or less 
With more than 2 rooms 

Electric Energy Production 

(In millions of k.w.h.) 620 914 

National Income 
(In IL millions) 

Private Consumption 
Expenditures 
(In IL millions) 

Manufacturing Employment 
(In thousands) 

Manufacturing Wages and 

Salaries (In IL millions) 198.1 269.0 

Families with Household Equipment 
(In thousands) 
With piano 
With gramophone 
With radio 

With electric washing machine ** 

With gas range or cooking stove 
With refrigerator— ice box 
With refrigerator— electric 

♦Average size of family 3.8. 

**In 1957, an investigation of 2,911 families revealed that 2.4% had bought an electric washing machine 
during the past year at an average price of IL. 329. Broken down by size of monthly income, the figures 
were 0.4% for 872 families with monthly incomes of less than IL. 200, 2.6% for 1,544 families with 
monthly income of IL. 200-349, and 6.2% for 417 families with monthly income of IL. 350 or more.$l = 
1.8 lirot (or "Israeli pounds," IL). 



1,726 2,399 2,721 3,047 

1,088 1,603 2,177 2,473 2,785 

81.7 - 99.3 

316.2 





15.4 




59.7 




384.3 


43.7 


61.6 


168.0 


239.5 


213.1 


216.3 


163.5 


204.8 



182 



PRODUCT POTENTIAL 



and a fan-type blower. An average washer 
load of clothes is damp-dryed for ironing 
in about 30 minutes and completely dried 
in about 45 minutes. The machine handles 
approximately 18 pounds of wet clothes 
at a time. A thermostat prevents the 
dryer from overheating and a selector con- 
trol can be set to time the operation for 
the desired degree of dryness. 



Major Market Factors 

1. Home Electrification. Since both gas 
and electric dryers require electricity to 
operate the tumbler and fan mechanisms, 
the basic requisite for ownership is a home 
with electric power. 

As of January 1, 1948 there were 33,- 
050,000 residential electric customers in 
the U.S. Of this number, 22,366,466 were 
urban customers, 6,912,504 were rural 
nonfarm customers; 3,771,000 were rural 
farm customers. 2 Thousands of new elec- 
trical customers, particularly in rural areas, 
are being added each year and by 1954 it 
is estimated that a total of 5,225,000 
farms will either have electricity or be 
within connecting distance of a hi-line. 3 

While it must not be presumed that 
every electrified home or farm represents 
a prospective buyer of an automatic clothes 
dryer, the mere fact that the potential 
market base is still expanding is, neverthe- 
less, an encouraging and promising devel- 
opment to the industry. Some idea of the 
market opportunities which exist may be 
observed from the fact that there were 
probably less than 100,000 automatic 
clothes dryers in use at the end of 1947, 
that is, that the "market saturation" of 



this appliance was no more than 0.3 per 
cent of all residential electric customers 
at that time. 

2. Climatic Conditions. Another im- 
portant market factor for the automatic 
clothes dryer is the fact that poor weather 
conditions often prevent housewives from 
hanging laundry out-of-doors to dry at a 
time of their own choosing. According to 
the latest U.S. Meteorological Year Book, 
there is cloudy or partly cloudy weather in 
180 American cities about 70 per cent of 
the time. This means an average of only 
2.07 clear days each week in these cities. 
Further checking on weather information, 
as supplied by the U.S. Weather Bureau, 
also shows that in most of the United 
States, with the exception of the Southern 
half of California and Nevada, most of 
Arizona and parts of New Mexico, Texas 
and Oklahoma, conditions of temperature, 
humidity, and a relatively low amount of 
sunshine definitely hinder good outdoor 
drying throughout the year. Certain areas, 
such as New England, the Middle Atlantic 
States, the Atlantic and Gulf Coast re- 
gions, the Pacific Northwest, and the 
Great Lakes regions have considerably 
worse drying weather than other sections. 
Consequently, the market for dryers in 
these places may have a somewhat higher 
sales potential than the average for the 
country as a whole. 4 

3. Impurity of Atmospheric Conditions. 
Contamination of the air from soot and 
dirt, being found most frequently in 
crowded industrial areas or near large fac- 
tories and railroads, increases the problem 
of home laundering when freshly washed 
clothes are hung outdoors to dry. The lo- 
cation of industrial centers is well-known, 



2 Source: Electrical Merchandising — Janu- 
ary 1, 1948. 

3 Source: Air Conditioning and Refrigeration 
News— May 24, 1948. 



4 Note that in Israel, in 1958, Tel Aviv had 
26 rainy days and 7 stormy days, while 
Jerusalem had 24 days of rain or snow and 
79 stormy days (Source: Statistical Abstract 
of Israel). 



PRODUCT POTENTIAL 



183 



but there are probably very few sections 
of the U.S. which are entirely free from 
soot and dirt in the air the year round. 
Because it tends to eliminate this hazard, 
the automatic clothes dryer should have 
a good potential market in areas of con- 
centrated industrial activity and in large 
cities. 

4. Availability of Drying Space. The 
frequent limitations of space for ade- 
quately drying clothes is another important 
element to be considered in estimating the 
market for the automatic dryer. This is 
particularly true in cities or urban areas 
where there is a heavy population concen- 
tration. In such areas where weather or 
atmospheric conditions are especially bad, 
a housewife must often resort to indoor 
drying, either in the basement or laundry 
room, which may not be large enough to 
accommodate a heavy washing. The auto- 
matic dryer thus possesses considerable 
sales appeal because it helps to solve dry- 
ing space problems. 

5. Traditional Attitudes of Consumer 
to Drying Methods. One of the most diffi- 
cult obstacles which the automatic dryer 
will have to overcome is the feeling on the 
part of many women that clothes dried in 
the sunlight or by other natural means 
would turn out better and have a sweeter 
smell than clothes dried by a mechanical 
process. The industry is thus confronted 
with the task of convincing prospective 
buyers that these traditional beliefs are not 
as valid as they have appeared to be and 
that clothes dried in an automatic dryer 
will be just as fresh as those dried in any 
other manner. This is mainly a problem 
of consumer education. Consistent and in- 
telligent advertising, directed at the con- 
sumer to point out the many advantages 
of the automatic dryer over other drying 
methods will be an important factor in 
overcoming this difficulty. 



6. Ability to Buy. The present high 
level of consumer income is a favorable 
omen for the automatic dryer market. It 
is impossible to predict, with absolute ac- 
curacy, the influence which the various 
levels of consumer income will exert in 
terms of actual sales. In all probability, 
however, the current price of the product 
and its comparative newness will result, 
in the early market stages, in developing 
the greatest volume of sales among the 
more prosperous families. As the advan- 
tages of the dryer become more widely 
known, as costs are brought down, and as 
terms of payment are adjusted to meet 
consumer requirements, the market should 
gradually broaden, thereby following the 
familiar pattern which has characterized 
the home appliance industry for years. 

7. Operating Costs. The cost of operat- 
ing an automatic clothes dryer appears to 
be quite reasonable and is a favorable 
market factor. Estimates made by the 
West Penn Power Company show that the 
electric dryer uses on the average about 
40 k.w.h. of electricity per month. At a 
rate of 3 cents per k.w.h. this amounts to 
$1.20 per month. Another estimate by 
Electrical Merchandising puts the cost at 
$11 per year. Since electric rates vary in 
different sections of the country, these 
cannot be taken as nationwide figures. 
Nevertheless, they do give some indication 
of the low cost of operation. . . . Costs of 
operation would also be affected to some 
extent by the size and frequency of home 
laundering. Many owners have no definite 
ideas of their dryer operating cost and do 
not seem to be concerned about it. It has 
been stated by dryer owners that they have 
not noticed any appreciable difference in 
their utility bills while others stress the 
fact that the convenience of the dryer 
more than offsets any increase in costs 
involved. 



184 



PRODUCT POTENTIAL 



Factors Determining Purchase 

Closely allied to the factors which will 
influence the growth and development of 
the automatic clothes dryer market are the 
respective advantages and disadvantages 
of the product and the reasons why people 
buy it. Dryer manufacturers cite a number 
of tangible advantages for their product, 
some of which are listed below: 

1. Reduces wash-day labor. 

2. Speeds up laundering operation. 

3. Gives clothes proper degree of damp- 
ness before ironing. 

4. Eliminates clotheslines and dangers of 
broken ropes. 

5. Eliminates dependence upon weather. 

6. Eliminates soot and dirt hazard in air. 

7. Heating temperature (140°-190°) sani- 
tizes clothes. 

Many of these advantages are supported 
by survey results. In a recent study of 500 
dryer users, conducted by a leading manu- 
facturer, 91 per cent of the respondents 
said they saved considerable time with a 
dryer; 81 per cent said its most important 
asset was the saving of work. 5 In an earlier 
issue, 6 Electrical Merchandising published 
an article giving records of time and effort 
saved before and after installation of an 
automatic dryer. In a typical instance it 
was found that the housewife walked an 
average of 0.8 of a mile for each drying 
operation before a dryer was installed; 
this walking was entirely eliminated after 
the purchase of a dryer. 

A number of clothes-dryer owners, 
mostly women, were also interviewed in 
the city of Philadelphia and vicinity by 
members of the Research Department for 



5 Source: Electrical Merchandising, May 1 
1948. 

6 March 1, 1947. 



the purpose of obtaining some expression 
of consumer opinion about the product 
and its performance. . . . The respondents 
in Philadelphia supplied additional verifi- 
cation of dryer advantages. Every single 
user of the appliance who expressed an 
opinion as to its value was pleased with its 
performance. Owners were particularly im- 
pressed with the softness and flufflness of 
items, especially Turkish towels, and at- 
tributed this to the tumbler action which 
keeps articles of clothing constantly in 
motion and does not permit them to 
stiffen as they dry. Others were pleased by 
the fact that many articles did not require 
ironing when removed from the dryer. 
Chenille bedspreads, blankets, small rugs, 
bath mats, men's underwear and children's 
playclothes were mentioned most often in 
this connection. 

Many favorable comments were record- 
ed on the number of steps saved, re- 
duction in the number of items that had 
to be ironed, freedom from wrinkling, and 
the elimination of sprinkling prior to iron- 
ing. Timesaving was another factor men- 
tioned quite frequently. Several stated that 
poor weather conditions were no longer a 
problem; the dryer insures good drying 
conditions regardless of the weather and 
permits them to wash on scheduled days — 
there is no backlog of dirty clothes when 
there is a dryer in the house. 

Most of the respondents are quite con- 
vinced that clothes come out of the dryer 
just as fresh as they do when dried in the 
sun; some even think it better than out- 
door drying because there is no wind to 
tear or fray the clothes, plus the fact that 
they are not as likely to get soiled from 
dirt in the air. Housewives also attached 
considerable importance to the fact that 
clothes seemed to keep their proper shape 
better in the dryer. 

Also noted were the reasons for buying 



PRODUCT POTENTIAL 



185 



an automatic dryer. These are arranged ac- 
cording to the frequency of mention as 
follows : 

Reasons for Buying a Dryer 

1. Lack of drying facilities in poor dry- 
ing weather. 

2. Inability or dislike to hang clothes in- 
doors. 

3. Desire to reduce laundry work. 

4. Desire to speed up laundry work. 

5. Desire to eliminate clothes soilage on 
outdoor line. 

6. Lack of adequate outdoor drying fa- 
cilities. 

7. Age or poor health. 

8. Good salesmanship by retailer. 

9. Gift. 

10. Need of special drying facilities for 
business. 

11. Clothes get torn in the wind. 

12. Heard about and decided to try one. 

13. Recommendations of friends. 
Certain disadvantages have also been 

claimed for the automatic clothes dryer. 
These are: 

1. Room Dampness. Because the dryer 
removes about 6 pounds of water from 
every load of clothes, the room in which 
it is located may become very damp while 
the dryer is in operation unless there is 
good ventilation. In some cases it has been 
found necessary to install a small exhaust 
fan to eliminate this moisture. 

2. Lint. Due to the tumbler action of 
the dryer, considerable amounts of lint are 
created in the drying process and lint 
traps must be attached to the machines. 
As these clog up when overfull, it is nec- 
essary to empty them periodically. 

3. Initial Cost. The present cost of the 
dryer ($220— $250) probably places it 
out of reach for many people. 

4. Fabric Damage. Certain fabrics may 



sometimes be damaged in the dryer due 
to high operating temperatures. . . . 

Some have said that the heat made silk 
things "harsh," while the majority of 
those who did not put woolen things in 
the dryer were afraid of shrinkage. Others 
said they had to be very careful not to 
leave the heat on too long, or the clothes 
would turn yellow. A few users stated that 
the room where the dryer was located gets 
very warm when the machine is running. 

The periodical Electrical Merchandis- 
ing for August, 1944, published the results 
of a door-to-door survey made during the 
war among owners of prewar clothes dry- 
ers. Among other things, it is interesting 
to note that some of the women owning 
the earlier model dryers had experienced 
trouble with burning or scorching. Many 
of the early defects and failures have un- 
doubtedly been eliminated in the post- 
war models. The women we talked to had 
never had any trouble of this kind, never 
heard of it happening to anyone, and 
were not worried in the least about it. 



Estimate of Potential Dryer Market 

For a number of years the principal 
market for the automatic clothes dryer will 
be new customer sales. Replacement and 
second-hand sales will not become an im- 
portant factor in the market picture until 
the appliance achieves a higher saturation 
status than it now has. An estimate of the 
potential market for the dryer, therefore, 
must primarily concern itself with a con- 
sideration of new sales possibilities. 

Since the drying of clothes is the next 
successive step after washing, there is 
good reason to believe that the pattern of 
market development for the clothes dryer 
will closely approximate that of . . . me- 
chanical home laundering in general. . . . 



186 



PRODUCT POTENTIAL 



In this connection it is important to ironers. These people contend that ironer 

note that some experts in the appliance owners are most conscious of the value of 

field have tied the potential dryer market, mechanical laundry aids and are also better 

exclusively, to ownership and sales of able to afford a dryer. . . . 

U.S. HOMES WITH ELECTRICAL APPLIANCES 
(number of appliances in millions) 

Wired Homes with: 

Vacuum Dryers 

Television Refrig- Cleaners Electric (Electric Air con- 

_ J Sets erators Freezers (Floor) Washers and Gas) ditioners 

End 

of Num- Per Num- Per Num.- Per Num- Per Num- Per Num- Per Num- Per 

Year ber cent her cent ber cent ber cent ber cent ber cent ber cent 

1946 - - 21.4 69.1 - - 15.1 48.8 18.8 60.5 - - 0.1 0.2 

1950 10.6 26.4 33.8 86.4 2.8 7.2 22.0 56.5 28.1 71.9 0.6 1.4 0.2 0.6 

1955 35.0 76.1 43.3 94.1 7.7 16.8 29.6 64.3 38; 7 84.1 4.2 9.2 2.6 5.6 

1958 44.0 89.0 48.3 97.7 10.4 21.0 35.0 70.9 44.9 90.9 7.7 15.6 5.8 11.7 

Note: Percentages based on total number of homes wired for electricity. 

Source: The Economic Almanac 1960 (National Industrial Conference Board), p. 371. 



Washing Machines 



SALES OF ELECTRIC APPLIANCES IN U.S. 
(number of appliances in thousands) 
Driers 



Year 


Automatic 


Other 


Electric 


Gat 


1936 


1,529 




- 


- 


1940 


1,455 




- 


- 


1945 


251' 


* 


251 


- 


1950 


1,646 


2,626 


68 


1955 


3,123 


1,268 


1,028 


369 


1958 


2,832 


938 


843 


397 


1959 


2,970 


980 


923 


486 


1960 


2,616 


765 


814 


444 



Combined 




Dish- 


Washer-Driers 


Ironers 


washers 


- 


180 


- 


- 


176 


- 


- 


30* 


- 


- 


409 


230 


- 


87 


295 


170 


35 


425 


196 


40 


547 


163 


35 


555 



♦Effects of war. 

Source: Statistical Abstract of the United States i960, p. 822; 1961, p. 821. 



PRODUCT POTENTIAL 



187 



CASE 3-7 

Lockheed Aircraft Corporation 



The practicality of long-range planning 
is a controversial question in industry to- 
day, and, in the planning area itself, one 
of the most disputed topics is long-range 
sales forecasting. 

How, businessmen may argue, can a 
firm foretell the future with certainty? The 
answer, in the opinion of a growing num- 
ber of executives, is that forecasts do not 
need to be certain. Then how, ask the 
critics, can long-range forecasting be of 
much help in planning? The answer is that 
long-range planning can be effective with- 
out dealing in certainties. 

According to this thinking, it is enough 
if management can work with probabili- 
ties — with the likelihood of certain poten- 
tial markets and certain market penetra- 
tions occurring under certain conditions, 
the chances of other conditions develop- 
ing which will affect the forecasts, the 
probabilities of various alternative devel- 
opments. In other words, management's 
aim should be to gain, not a hard and fast 
outline of the future, but an evaluation of 
probabilities on which it can make in- 
formed decisions. 

This is the thinking that has animated 
an approach developed at Lockheed Air- 
craft Corporation. We have used the 
method, dubbed "prudent-manager fore- 
casting," with some success for a number 
of years. We believe that many other com- 
panies could also use it with profit, es- 
pecially firms selling products and services 
to markets characterized by relatively 
small numbers of large buyers and sellers. 7 



7 This description is by Gerald A. Busch, 
Director of Market Research at Lockheed 
Aircraft Corporation. See "Prudent-Manager 
Forecasting," Harvard Business Review, 
May-June 1961, p. 57. 



To apply this technique, management 
(a) brings together a small group of sea- 
soned specialists representing such func- 
tions as marketing research, marketing, 
finance, engineering, and administration, 
and (b) asks them to assume the role of 
decision-making managers in a customer 
firm that is evaluating one of the firm's 
products for purchase. In effect, this group 
of specialists assumes the position of the 
customer's management. In so doing, it 
attempts to evaluate prudently the facts 
available and to arrive at the preferred 
procurement decisions — that is, preferred 
from the customer's point of view. 

In a way, this is very much like the 
"role playing" used in management train- 
ing, in that company people put on the 
hat of somebody else and assumes that 
"somebody else's" point of view. The 
origin of the idea probably lies in the 
sociodrama that has been used with some 
success by clinical and group psychologists 
in ironing out family problems. 



PREPARATORY STEPS 

But prudent-manager forecasting can- 
not take place in a vacuum. It will not 
work unless it is preceded by a great deal 
of staff work. To be successful, those who 
are asked to serve as prudent managers 
must be provided with adequate, carefully 
organized information to draw on in the 
decision-making process. They should not 
be expected to make decisions without 
having the facts necessary for rational 
judgment. This is where the work of the 
market-research staff comes in. 

Since the preliminary, analytical work 
is so vital to the successful application of 



188 



PRODUCT POTENTIAL 



the prudent-manager technique, I shall be- 
gin with a brief discussion of some of the 
major stages of preparation. Note that the 
steps described are actually those involved 
in the classic analytical approach to fore- 
casting, and in themselves will provide a 
forecast that may be sufficient for certain 
purposes. However, when top management 
requires a longer term sales forecast, the 
addition of the prudent-manager technique 
is far more than frosting on the cake. As 
we shall see, it adds substance, acumen, 
and depth of viewpoint to management's 
expectations. 

The main steps in forecasting can be 
illustrated from Lockheed's experience. 
Executives of other companies should not 
find it difficult to visualize data from their 
own studies in place of the specifics shown 
here. 



Ground Rules 

It is typically necessary to narrow the 
scope of a study so that it will be com- 
patible with the elapsed time, manpower, 
and other resource constraints surrounding 
the study. Accordingly, we start with cer- 
tain ground rules which are tailored to the 
specific forecasting problem at hand. For 
example, in a study in 1959 having to do 
with our electronics sales outlook, we 
set forth three categories of ground rules 
and assumptions — economic, political, and 
technological (see Figure 3-J) . In each case 
we drew on the best thinking of both line 
and staff executives. To illustrate from 
Figure 3-J: 

(1) Some of the points were derived 
from previous studies focusing on the sub- 
ject at hand. The last four items in the 
"economic" group are a case in point. 
Here we dealt with projections of the 
gross national product, total major na- 
tional security expenditures, and certain 
components of the latter — for instance, 



Department of Defense procurements and 
procurement by the National Aeronautics 
and Space Administration. These were all 
areas in our business warranting continued 
investigation. (In the case of a company 
selling, let us say, bearings to large manu- 
facturers of diesel engines, a similar state- 
ment might be made about projections of 
the number of end users, transportation 
trends, gross national product, and related 
factors.) 

(2) Some assumptions were based sim- 
ply on informed judgment. For instance, 
our "political" assumptions were that 
there would be no hot war and no sub- 
stantial disarmament. These appraisals 
may, of course, turn out to be completely 
wrong. And yet we had to make assump- 
tions to begin with, and these seemed like 
the most reasonable ones. 

(3) Some assumptions were based on 
specialized technical knowledge. Examples 
here are the "technological" estimates that 
the improvement in missiles would be 
evolutionary rather than revolutionary, 
and that a five-year lead time would be 
needed for commercial application. (In 
the case of the bearings manufacturer, 
comparable projections might concern the 
development of new materials which would 
affect the use of bearings.) 

You may quarrel with individual items 
in this list of ground rules; and, of course, 
we do not recommend them for other 
companies' programs. The point is that in 
long-term sales forecasting you generally 
cannot afford the luxury of a "womb to 
tomb" approach, so limiting ground rules 
are necessary. Management should make a 
practice of stating these explicitly. 



Characteristics of Demand 

The next stage requires a look at the 
historical characteristics of supply and de- 
mand in the markets that management is 




Figure 3-J. Major Ground Rules and Assumptions 





TOTAL (BILLIONS OF DOLLARS) 




1951- 


$1.92 


1954- 


$6.75 


1957- 


$6.14 


1952- 


- 3.74 


1955- 


5.98 


1958- 


6.24 


1953 


- 6.12 


1956- 


5.59 


1959- 


4.97 










\ 


\ y 




TRAINERS AND 
OTHER AIRCRAFT 



Figure 3-K. Historical Characteristics of Demand for Manned Aircraft by the U. S. Air Force 



189 



190 



PRODUCT POTENTIAL 



concerned with. Ordinarily, this means 
studies not only of demand for different 
products but also of the buying behavior 
of different customers. At Lockheed, for 
example, we are deeply interested in 
such customers as the military services, 
other government agencies, and certain 
commercial buyers; and in the demand 
characteristics for aircraft, missiles, space 



vehicles, electronics, and ground-handling 
equipment, as well as a host of other 
products. 

Figure 3-K shows a study we made last 
year in one area of importance to us — Air 
Force spending of its aircraft procurement 
budget. This analysis gives us a picture of 
how these funds were expended during the 
decade of the 1950's — for bombers, fight- 



INTERC1TY CARGO 

TRANSPORTATION DEMAND 

(TON MILES) 




Figure 3-L. Correlation Between U. S. Domestic Cargo Transportation Demand and GNP 



PRODUCT POTENTIAL 



191 



ers, transports, trainers, and other types of 
aircraft, as well as for spare parts, ground- 
handling equipment, and modification. 
From this type of study we can make 
comparisons and correlations with con- 
current operational, technological, politi- 
cal, and economic events, and then develop 
appropriate forecast indicators. There is 
reason to believe in the value to the 
forecaster of a good hard look at history. 

Economic Indicators 

Next comes the development of eco- 
nomic indicators. Correlation analysis may 
be very helpful here. A case in point is a 
Lockheed study of the relation between 
cargo transportation and gross national 
product. Figure 3-L portrays: (1) the ef- 
fective demand for cargo transportation 
service in the U.S. domestic market since 
1929 as compared to (2) the gross na- 
tional product in terms of constant dol- 
lars. 

The correlation between these two series 
is 0.97 — high enough to permit us to 
project the demand for cargo transporta- 
tion on the basis of our forecasts of gross 
national product. We at Lockheed are in- 
terested in the total cargo picture, of 
course, because the total demand is rele- 
vant to the demand for air cargo, in which 
we are more directly concerned. 

It should be added that it is seldom we 
get such a high correlation as in Figure 
3-L. As forecasters know, the closer to the 
actual product an analysis gets, the lower 
is the correlation usually obtained. 

The next step is to derive a parameter- 
ized market projection. In such an analysis, 
the aim is to look at the effect on the 
market of changes of the key variables. 
Briefly, the procedure is as follows: 

( 1 ) For the problem at hand, the fore- 
casters select the significant variables — 
like consumer income, price and availa- 



bility of service, system performance — and 
decide what the "most likely" future lev- 
els will be. 

(2) Then, assuming that these "most 
likely" levels will turn out to be correct, 
they make an estimate of demand for the 
product or service. 

(3) But they must make allowances for 
other possibilities; the "most likely" lev- 
els may not in fact be realized. Therefore, 
they make projections of demand based 
on other levels which may reasonably 
occur. 

(4) When they have a range of de- 
mand or "output" figures correlated with 
different "input" levels for one variable, 
they take another variable and repeat the 
process. 

(5) The final result is a matrix of pro- 
jections showing how sensitive the fore- 
cast is to various changes in each of the 
input variables. 

Such a "sensitivity analysis," as it is 
called, is valuable to the prudent man- 
agers when, subsequently, they deliberate 
over the forecast. It is likely to show, for 
instance, that certain variables are more 
important to watch than others. To use 
an illustration again from the experience 
of Lockheed: Figure 3-M shows a "para- 
meterized" forecast made a few years ago. 
At the left is the historical trend in the 
distribution of the demand for air, rail, 
and bus transportation in the United 
States. At the right are shown five alter- 
native projections: 

In Projection A all of the inputs are 
at their "most likely" projected levels. 
Projection B shows what demand would 
look like if we held all other inputs con- 
stant, but restricted the availability of air 
coach service to no more than what it was 
in 1953. In Projection C we have re- 
stricted the performance of the surface 
systems — bus and rail — to their 1953 lev- 
els. In Projection D we have increased the 



192 



PRODUCT POTENTIAL 



HISTORICAL DISTRIBUTION 



ALTERNATIVE FORECASTS OF 1965 
DISTRIBUTION, - 




1920 1930 1940 1945 1950 1953 A B C D 

Figure 3-M. Distribution of Demand for Bus, Rail, and Air Transportation 



unit valuation on the time of the trav- 
eler by 50 per cent over the projected lev- 
els. And in Projection E we show what 
the effect on the picture is in the event 
that the rate of expansion of the economy 
of the United States should turn out to 
be somewhat less than our other studies 
suggest. 

Thus we see at a glance how variations 
in the input variables will affect the total 
demand for air passenger transportation 
in 1965. Incidentally, the staff did not 
make firm conclusions or choices among 
the several situations. We concluded that 
the demand for airline transportation 
would increase between threefold and 
fivefold from 1953 to the early 1970's and 
that, assuming all inputs were at their 
"most likely" projected levels, the poten- 



tial air transportation demand would be 
about 57 billion great circle passenger 
miles per year at that latter time. But we 
pointed out that this conclusion was 
highly sensitive to two of the inputs — 
(1) the projected expansion of the U.S. 
economy, and (2) the projected perform- 
ance of the surface transportation systems. 
(Note that air transport is highest in Pro- 
jection C, where it is assumed that per- 
formance of surface transport systems will 
stay at the same level as in 1953.) 



COMPARISONS WITH COMPETITORS 

The next step is to take an objective 
look at the company's product or service 
in comparison with its competition. It is 



PRODUCT POTENTIAL 193 



important to select those characteristics 
for comparison which are of particular 
concern to the potential buyer. To illus- 
trate, Figure 3-N shows a few of the char- 
acteristics of competing transport-type air- 
craft of concern to Lockheed's customers. 
Here we see comparative data on speed 
performance, pay-load range performance, 
unit operating-cost performance, and re- 
turn on investment. With comparative 
data such as these, the prudent manager 
can subsequently make judgments as to 
the relative importance in the eyes of the 
customer of, say, differences in block 
speed and return on invested capital. 

Rounding Out the Forecast 

Now, with all this work done, the mar- 
ket-research staff is able to make a fore- 
cast of sales of the Lockheed product — 
again preferably in parameterized form. 
This it does on a customer-by-customer 
basis, if feasible. The procedure is as fol- 
lows: 

(1) The major customer categories are 
separately analyzed. For instance, in mak- 
ing a study of the sales outlook for a cer- 
tain type of electronic end-equipment, 
the Lockheed staff looked at prospective 
sales to the government, including the 
Air Force, Navy, Army, and NASA, and 
to the nongovernment agencies. 

(2) Within each customer category, 
the sales in each application are separately 
forecast, as in Figure 3-0. (To simplify this 
chart, the data are shown in nonparameter- 
ized form.) 

(3) The staff puts all of the individual 
forecasts together and comes up with an 
aggregate sales forecast. 

(4) The aggregate figure is subjected to 
a test of reasonableness. After adding up 



Figure 3-N. Comparative Performance Char- 
acteristics of Cargo Transport Aircraft 




194 



PRODUCT POTENTIAL 




Figure 3-0. Estimated Procurement of an Item of 
Electronic End Equipment for Different Applications 



all the individual forecasts, the staff asks 
itself, "Does the resulting total appear 
reasonable? Is it in consonance with our 
traditional percentage of the market? 
Does it appear too pessimistic or too op- 
timistic? Is too much weight given to 
some factors and too little to others?" 
Only after the forecast has been subjected 
to questions like these and, if necessary, 
revised, is the stage set for the prudent 
manager. 

MORE REALISTIC JUDGMENTS 

All the steps and techniques described 
up to this point are part of the staff 
work of forecasting in many compa- 
nies. At Lockheed, some details may be 
changed or emphasized more in anticipa- 
tion of what comes later, but in the main 
the type of work just outlined is essential 
to any intelligent kind of forecasting, and 
even by itself would lay the groundwork 
for informed maangerial decisions. 

However, it does have the disadvan- 
tage of inbreeding. It is done by staff 
people who sometimes may become be- 
mused with figures and, in their concen- 
tration on the tools and techniques of 
their craft, lose contact with the reality 



with which they should be dealing and 
which, in effect, they are trying to fore- 
cast. 

Therefore, to reintroduce this factor of 
outside reality and to help assure that our 
sales forecast will give service to the com- 
manding importance of the customer's 
point of view, we like to expose the staff 
analysis to a balanced group made up pre- 
dominantly of non-marketing research peo- 
ple, and get their evaluation. This brings 
us to the essence of the prudent-manager 
approach. 



Organizing the Conference 

To begin, there is the question of 
how these non-marketing-research people 
should be organized. Our experience at 
Lockheed indicates that it is a good idea 
to try to make up a conference from a 
representative group, but we have found 
it essential to keep the number of par- 
ticipants small. 

So far as it is practical to do so, we 
use Lockheed people. We will usually in- 
clude a representative from finance who 
knows firsthand the financial situation of 
our customers, one or two representatives 
from marketing who are very knowledge- 
able of the market and the customers in 
question, and one or two managers from 
the technical side of the house who have 
a detailed knowledge of the product's per- 
formance. Occasionally, we include an 
outside consultant — an economist from a 
university, perhaps, or a business consult- 
ant. And we always include a senior mar- 
ket-research staff member who is familiar 
with all of the preparatory work done by 
the staff. 

In general, the participants should be 
mature, seasoned in their areas of spe- 
cialization, objective in their approach to 
problems, and able to both give and take 
in the negotiating process. 



PRODUCT POTENTIAL 



195 



We have found that it is desirable, 
prior to the first get-together of the pru- 
dent managers, to provide each man with 
copies of the preliminary staff reports in 
order that he may do some "homework" 
on the subject beforehand. 

At the first formal meeting, the pru- 
dent managers are briefed in detail on the 
staff work. Sometimes we devote as much 
as one day to this briefing so that the par- 
ticipants will know in detail how each 
input was developed, how the variables 
were integrated, and so forth. From 
there on, the group is on its own. Note 
the importance of adequately briefing the 
prudent managers — for if this is not done 
well, there is the distinct danger that the 
group will resort to that insidious tech- 
nique which can be called "feet-on-the- 
deskmanship." 



Reaching Conclusions 

Now let us turn to the deliberative proc- 
ess itself. For the sake of being specific, 
let us suppose that the managers are con- 
sidering the problem of arriving at a long- 
term sales forecast for a proposed new pas- 
senger transport airplane. How do they 
go about making their judgments? 

To begin with, they put themselves in 
the place of each major potential custo- 
mer, in turn, starting with the larger air- 
lines and later dealing with relatively ho- 
mogeneous groups of small airlines. In 
this position, and generally accepting the 
analytical work done by the staff as "gos- 
pel," they carefully review the judgments 
made by the staff — and typically interject 
others which the staff may have over- 
looked. 

To be more specific still, let us suppose 
that the customer under consideration is 
Air France, a major foreign airline. Each 
participant figuratively stands in the shoes 
of the manager of Air France and at- 



tempts to answer questions such as these: 

1. What new aircraft procurement policy 
must I pursue in order that my airline 
will remain competitive and retain or 
improve its standing in the market over 
the next decade? 

2. How does the proposed new Lockheed 
plane stack up against competitive of- 
ferings? 

3. Will I be able to finance my new air- 
craft? 

4. How many planes should I buy, say, 
in 1965? 

5. What will be my needs over short-haul 
routes, long-haul routes, and so on? 

6. How much money can I make with 
the Lockheed aircraft as compared to 
what I can make with competitive 
planes? 

7. What about return on investment? 

8. What about operations into marginal 
airports? 

In pursuing this approach, each par- 
ticipant tries to appraise dispassionately 
the Lockheed product vis-a-vis competing 
products in each customer's application. 
The prudent managers must be hard- 
headed about this — just as hardheaded, 
we hope, as they would be if they were 
charged with the responsibility of actually 
running the company they are consider- 
ing. Very often they know something 
about the personal failings and strengths, 
the preferences and prejudices, of the men 
whose positions they are assuming. For 
example, Mr. A likes pod-mounted en- 
gines, while Mr. B has been publicly 
thumping the drums for high-density 
coach configurations. All these items are 
taken into account as they argue out and 
discuss the probable buy the customer will 
make. 

Next, the prudent managers work out 
a forecast of that customer's likely pur- 
chases of the Lockheed airplane. Then, 
having settled on what they think is the 



196 



PRODUCT POTENTIAL 



most probable purchase of one customer, 
they pass on to another and start the 
whole process over again. 

When they have covered the whole 
field of potential customers, they total the 
probable purchases and come up with an 
aggregate forecast. This result is then 
compared with the staff results and, as in 
the staff procedure, is subjected to the 
general test of reasonableness. 

The result, we think, gives a closer ap- 
proximation to what may in fact actually 
happen than can mere staff work alone. 



TESTS OF EFFECTIVENESS 

How effective prudent-manager sessions 
are depends in part on the thoroughness 
of the staff work done in preparation for 
them, for a lot of the managers' thinking 
is based on this work. But effectiveness 
also depends on how well the talents and 
viewpoints represented mix together. 

Generally speaking, the level of compe- 
tence increases as the organizational level 
of the members of the team increases, be- 
cause the men higher in the organiza- 
tion are generally those with the greatest 
knowledge and acumen, and the broad- 
est point of view. But there are drawbacks 
to inviting men who sit too high in the 
organization. For one thing, they may not 
find it possible to stay with the group be- 
cause of other demands on their time. For 
another, a vice president sitting with men 
two levels below may unknowingly tend 
to constrain valuable give-and-take discus- 
sion. 

Beyond this, we have found that there 
are some people who have a natural tal- 
ent for this sort of discussion. They have 
the imagination to put themselves in 
someone else's place, the ability to judge 
dispassionately and realistically, and the 
breadth to consider all phases of the prob- 
lem. Generally speaking, we have found 



that people in senior staff jobs in the 
company measure up best in this respect. 



CONCLUSION 

To sum up, certain lessons based on 
experience in using the prudent-manager 
techniques should be emphasized. 

In organizing a conference, these are 
the most important points for manage- 
ment to remember: 

(1) Keep the number of participants 
small. 

(2) Include a balanced representation 
of technical specialties. Bring in (a) se- 
lected "in house" specialists, (b) a senior 
market-research staff analyst, and (c) an 
occasional outside consultant, as needed. 

(3) Give the participants a detailed 
briefing on the background and results of 
the staff study. 

To make the deliberative process a suc- 
cess, each prudent manager should: 

(1) Assume the position of decision- 
making executives in the customer firm. 

(2) Dispassionately appraise the per- 
formance of competing products that the 
customer firm might consider buying. 

(3) Make a forecast of the purchases 
that he thinks the customer will make. 

(4) When the individual forecasts are 
pulled together into an aggregate fore- 
cast, join with the other participants in 
testing its reasonableness. 

Keep in mind throughout all this that 
prudent-manager sessions are not a sub- 
stitute for sound staff forecasting; they are 
merely a refinement on it. Certainly they 
should never be used instead of such staff 
work. Management should also be fore- 
warned that the technique is relatively 
costly — particularly in the early years of 
its application. As a rule, at Lockheed, we 
"roll over" our important long-term fore- 
casts every year or two. The time and 
effort required in the second and subse- 



PRODUCT POTENTIAL 



197 



quent passes is significantly less than in 
the original forecast. 

It is not clear that prudent-manager 
sessions are the answer to all forecasting 
problems or, indeed, that they should 
even be applied to all such problems. But 
it is clear that, where the sales fore- 
cast is of particular importance to top 
management in its long-range planning, 
and where objectivity and realism are pri- 
mary goals, this technique may be a use- 
ful addition to present forecasting meth- 
ods. There is every reason to believe that 



the benefits gained at Lockheed can be 
duplicated at many other companies. 



How useful is this approach likely to be 
as a basis for making decisions about new 
products — product characteristics, pricing, 
sales plans, and so forth? Is it applica- 
ble to industrial products like grinding 
wheels, electrical motors, or control in- 
struments? To consumer products? Or 
could it be adapted to such uses? 






198 



PRODUCT POTENTIAL 



In 1956, the Dow Chemical Company, 
with sales of 565 million dollars and net 
income of 60 million dollars, was a ma- 
jor producer of industrial chemicals, plas- 
tics, agricultural chemicals and magne- 
sium. In recent years, the company had 
also entered the field of consumer prod- 
ucts — most notably with its highly suc- 
cessful Saran Wrap. 

In 1950, Dow had organized a two-man 
market research department. By 1956, this 
department had grown until its personnel 
included a manager, an assistant man- 
ager, and 10 specialists in the company's 
several product lines. 

One of the early projects assigned the 
market research department in 1951 had 
been the evaluation of the market for 
slag wool. 

Approximately 21,000 tons of glasslike 
ash were being produced as residue by the 
company's Midland, Michigan, power 
house each year. The ash presented a dis- 
posal problem which was certain to grow 
as the power plant expanded to service 
the growing chemical works. Analysis of 
the ash indicated that it had the same 
general properties as materials used in 
making rock wool. By adding the one step 
of forcing the hot liquid slag through a 
small orifice and subjecting it to a jet of 
steam, mineral wool, useful for insulation, 
could be produced. 

Dow was about to install a second unit 
of its power plant. To modify the furnace 
of this unit to facilitate the rapid con- 
veyance of liquid slag would require an 
incremental investment of approximately 
$100,000. It was estimated that a plant 
for producing rock wool would cost an 
additional $250,000. 

The engineering department felt that 



CASE 3-8 

Dow Chemical Company (B) 

profits from rock wool might prove to be 
a welcome offset to the cost of power. 
The market research department was 
asked to investigate the possibilities. Its 
report follows: 

INTRODUCTION 

The engineering department, by report 
on June 11, 1951, has outlined the de- 
sirability of finding useful, salable prod- 
ucts which can be made from molten slag 
produced by the cyclone type boilers in 
the South Power House. The present two 
boilers produce molten slag at the rate of 
30 to 40 tons per boiler per day or ap- 
proximately 21,000 tons for all boilers 
per year. The slag disposal system for the 
South Power House cost $424,000 to in- 
stall and is continually incurring opera- 
tional expense. The planned addition of 
two new boilers of similar type has stim- 
ulated interest in developing uses for the 
slag which would provide a return on 
the equipment installation and alleviate 
the physical problem of slag disposal. 

Products which have been considered 
as possible outlets for the slag are light- 
weight aggregate for building blocks and 
slag wool for residential and industrial in- 
sulation. The chemical composition of 
the slag is not entirely satisfactory for 
building blocks due to high iron concen- 
tration. No wool has yet been made from 
our slag to determine the quality of prod- 
uct which could be produced but the 
wide variety of raw materials which have 
been successfully used by mineral wool 
manufacturers indicates that production 
of wool from our boiler slags is well within 
the realm of possibility. 

Market research has been requested to 



PRODUCT POTENTIAL 



199 



investigate the nature of the mineral wool 
industry and the market considerations of 
making and distributing a slag wool. 



A manufacturing price of $30 to $35 
per ton, based on loose wool, should be 
attractive to such a distributor. 



Findings 

Although no wool has yet been made 
from our cyclone boiler slag, there is a 
good possibility that a satisfactory prod- 
uct can be produced, considering the va- 
riety of raw materials used by existing 
mineral wool plants. Additional compo- 
nents may have to be added to the slag to 
improve fiber characteristics. 

The mineral wool industry is charac- 
terized by high material and labor costs 
(80 per cent of manufacturing price) 
and we would enjoy a competitive ad- 
vantage over some plants by using a par- 
tially processed waste material. 

Slag from two boilers would produce 
around 16,000 tons of loose wool per year. 
It is estimated that an area containing 
1,600,000 people could support mineral 
wool production of 10,000 tons per year. 
Michigan has 6,300,000 people, of which 
more than half reside in the southern 
part of the state. This population would 
require around 39,000 tons of mineral 
wool per year. 

Mineral wool is uneconomical to ship 
more than 300-400 miles. Our market 
area would therefore be Michigan, Ohio, 
Illinois, Indiana, and Wisconsin. This 
area contains the highest concentration 
of mineral wool plants in the United 
States. 

Distribution of product is the most im- 
portant factor in successful mineral wool 
operations. If a distributor could be 
found who has no plant in our logical 
market area, or who is desirous of sup- 
plementing a line of allied products, the 
project should be attractive to us as a 
means of returning something on equip- 
ment investment and disposing of waste 
materials. 



THE PRODUCT 

Mineral wool is a substance composed 
of very fine, interlaced, mineral fibers 
having the appearance of loose wool or 
cotton. This fibrous, glasslike, material is 
composed principally of silicates of cal- 
cium and aluminum together with other 
minor constituents. 

Mineral wool is a generic term cover- 
ing a number of similar products dif- 
ferentiated chiefly by the raw materials 
from which they are made, as follows: 

Rock Wool, made from natural rock 
or from various combinations of natural 
minerals. 

Slag Wool, made from iron, copper, or 
lead blast furnace slag alone or mixed 
with fluxing materials. 

Glass Wool, made from silica sand, 
soda ash, and limestone with or without 
scrap glass or other materials. 

The marketable forms of mineral wool 
most generally encountered are: 

Loose Wool. The raw product as blown 
and without subsequent treatment. This 
is the material from which all mineral 
wool products are made. It can be used 
as insulation in new houses without fur- 
ther treatment. 

Granulated Wool. Loose wool is run 
through granulators that break up the fi- 
bers and free them from "shot." Freed 
from shot and rolled into small pellets Yi 
inch or less in diameter, the wool is 
known and sold as granulated wool. The 
principal use in this form is as insulation 
for houses already built. 

Fabricated Forms 

Batts. Loose wool mixed with suitable 
binders is compressed and cut to such 



200 



PRODUCT POTENTIAL 



thickness and shape as consumers may 
demand. In this form it is used as insula- 
tion in new houses by insertion in walls, 
floors or roofs between studs, joists, etc. 
Batts are supplied with or without one or 
two paper covers. 

Blankets. Loose wool, mixed with suit- 
able binders, is formed into blankets of 
various rectangular shapes and thickness 
and sewed between paper, chicken wire 
or other confining media. 

Blocks. Loose wool with suitable bind- 
ers is compressed into blocks for use in 
refrigerator or stove insulation. 

Pipe Covering. Loose wool can be 
compressed and formed to produce pipe 
covering. 

Insulating Cement. The loose wool can 
be combined with clay, bentonite, and as- 
bestos fibers and sold as an insulating ce- 
ment. 

Raw materials for mineral wool produc- 
tion include a wide variety of suitable 
rock formations and slags. Most mineral 
wool is now made from slag. An example 
of a slag composition used by a mineral 
wool producer operating on copper slag 
is similar to a typical analysis of our cy- 
clone boiler slag: 





Maximum 


Mi 


nimum 




% 




% 


Si0 2 


38 




27 


FeO 


40 




35 


A1 2 3 


14 




3 


CaO 


1.5 




.4 



The Bureau of Mines reports the fol- 
lowing average recovery from 1,000 pounds 
of raw materials, based on the experience 
of several plants: 



This report will deal largely with min- 
eral wool which is produced for sale only 
as residential insulation and which ac- 
counts for somewhat more than half the 
total mineral wool market. The forms of 
product will therefore consist of loose 
wool, granulated wool, and batts and 
blankets. 



THE PRODUCERS 

The Bureau of Census, 1947 Census of 
Manufactures, lists 91 establishments em- 
ploying 8,874 workers engaged in all 
types of mineral wool production. The 
National Mineral Wool Association, whose 
members account for 80 per cent of resi- 
dential mineral wool production, has 
listed 75 member and nonmember plants. 
Figure 3-P shows the producing compa- 
nies and the location of mineral wool 
plants. As can be seen from the chart, 
there is a heavy concentration of plants 
in the East Central and West Central 
states especially in Indiana, Illinois, and 
Wisconsin. New Jersey also has numerous 
plants operated by the larger companies. 
Michigan has only the one plant at Kala- 
mazoo operated by Baldwin-Hill, but is 
completely encircled by numerous plants 
in neighboring states. 

Figure 3-Q illustrates the various types 
of mineral wool products made by a rep- 
resentative group of manufacturers and 
the raw materials used by these compa- 
nies. Nearly all companies make the full 
line of products. 

In addition to the companies with pro- 
ducing plants, the companies listed on the 
next page distribute a line of mineral wool 



From 

Lead Slag 
Iron Slag 



Loose Wool 
Pounds 

760 
880 



Granulated Wool 
Pounds 

500 
620 



Shot Loss 
Pounds 

260 
260 



PRODUCT POTENTIAL 



201 



The Flintkote Company 
Insulite Division 

Minnesota & Ontario Paper Co. 
The Lehon Company 
Armstrong Cork Company 
Barrett Division 

Allied Chemical & Dye 
The Celotex Corp. 

(Have only one plant and buy 

extensively from other producers) 
United States Gypsum Company 

products along with their other insulation 
or allied materials. 

In 1939 the United States Bureau of 
Mines reported an annual United States 
capacity of 600,000 tons of loose wool 
which was 50 per cent above estimated 
production for 1938. In 1947 the Census 
of Manufactures reported shipments and 
interplant transfers of 960,000 tons val- 
ued at $79,094,000 f.o.b. plant. An esti- 
mate of the 1950 production of mineral 
wool for residential building applications 
would total 1,200,000 tons with a value 
of $45 million to $50 million. The Na- 
tional Mineral Wool Association esti- 
mates that present capacity for wool pro- 
duction exceeds current demand and that 
substantial increases in output could be 
secured if defense requirements are raised. 

The 1947 Census of Manufactures re- 
veals the following operational data on 
mineral wool plants: 



New York, N.Y. 

Minneapolis, Minn. 
Chicago, 111. 
Lancaster, Pa. 

New York, N.Y. 
Chicago, 111. 



Chicago, 111. 



of a mineral wool plant and that the oper- 
ating margin and net profit, based on the 
above estimates, are small. 

In our own consideration of slag wool 
manufacture, an obvious advantage would 
be the ability to utilize a waste product 
which had been partially processed, thus 
eliminating certain of the material and 
labor charges. It may be necessary to add 
other components, such as calcium-bear- 
ing materials, to our slag in order to pro- 
duce wool having a composition required 
by the trade. Mineral wool companies 
state that a high iron content slag has 
excellent fluxing characteristics but needs 
other additives to improve fiber qualities. 
Wool made from slag high in iron is 
dark in color, but this is not too important 
since the most popular items, batts and 
blankets, contain a phenolic resin binder 
which imparts a brown color even to 
white fibers. 









% of value 








of shipments 


Value of shipments from plant 




$73,131,000 


100% 


Cost of materials, fuel, etc. 


$36,291,000 




49 


Wages and Salaries 


22,924,000 




31 


Total 




59,215,000 




Gross Profit 




$13,916,000 


20 


Estimated Selling & Adm. Expenses 




5,800,000 


8 


Profit Before Taxes 




$ 8,116,000 


11 


Estimated Taxes in 1947 




3,000,000 




Net Profit 




$ 5,116,000 


7 



These figures are only representative 
and such expenses as depreciation, income 
charges, and so forth, are not considered. 
They do reveal, however, that materials 
and wages are a large portion of the costs 



In the type of process contemplated for 
utilization of our boiler slag for wool pro- 
duction, careful consideration would have 
to be given to the implications of large- 
scale manufacture. 



la 







&.*<£ I 



I £ !l £ <5 



5 „: 



I * <£ J <3 -3 | J 3 ; C 



g g «S E 1 = -« rf 



c ;: J= * — 



1 iJlJJ^JI-El^ 
S J | 2 f 1 1 ? | * | 

J 3 3 I « S S S S S i: 



2 I i = g I S J s S "I 1 s | | g 

III I 3 -I | ° I ill ill 4 

1 5 I * I 3 i 1 3 1 1 i 1 1 j | 

iJlllllllJI^Jl 



< d 



•5 ■= 4? S - - 

-. * | | | | 5 j 5 i i i 



3 <3 £ J j * - 2 | J IS | | £ | 1 1 






^ ^?sS^*sj*5iK!3EKKS 



INIIitiiS 

3 

I 



■* jj5| mi n jfi 

II 3 ! : : : =i «= 1 * J 



i jd 



nit ^ 

i 1 1 £ d £ 



j I d d 6 

1 * 1 1 1 * I | <3 £ 1 I 1 J 

I Mill 11 £11^1 

8 j E E E £ i 3 C | J I I I 

|l I ! si il> 11?11 

ai»^Ri«Rsi5is»si*:^*;B5 



333 333 

8 11111 



11111 liJi Hi ^ ? ~ 1 1 = 
||1111f!f j3f|l||ll| 

1 1 1 1 1 1 1 II j | j 1 1 1 II 3 3 



f»i -i ^ ifl <i (^ * »' s 



202 



PRODUCT POTENTIAL 



203 



PRODUCT 
DISTRIBUTION - 
20 COMPANIES 


MAT! 


BATTS 


BLANKETS 


LtfG 


< 

-I 


o 
o 

Or 


in 
< 



i&j 
o 

<7i 

CQ 
> 


Q 
bJ 



—1 
O 

z 

tl_J 


*- ^ (9 


CO 


Q 
to 

CQ 
> 


Q 

L-J 

o 

_J 

u 




CN 


* 


* 

CO 


V 


UJ 

o 

O 

_i 


o 

U-l 

z 

< 

o 


AMERICAN ROCK WOOL 


▼ 






• 


• 




»• 




• 


• 




* 


* 






▼ 


▼ 


BALDWIN-HH.L 


T 






• 


• 








• 


• 




* 


• 






▼ 


▼ 


CARNEY CO. 


T 






• 


• 








* 


• 




* 


• 






▼ 


▼ 


CELOTEX 


T 








• 










• 




• 


• 






▼ 




COLORADO INSULATING 


T 






• 




• 4 




• 


















EAGLE-PICHER 


T 






• 
























▼ 


▼ 


FEDERAL PORTLAND CEMENT 


T 






• 


• 






• 
















▼ 


▼ 


FELTROK INSULATION 


▼ 






• 


• 






• 


* 


• 




• 


• 


* 




▼ 


▼ 


JOHNS-MANVILLE 


T 






• 






k tm< 


• 




• 




* 








▼ 


▼ 


MINERAL WOOL INSULATIONS 


T 






• 








• 


* 






• 










▼ 


MORELL 


T 






























▼ 


▼ 


MUNFORD COMPANY 


T 






























▼ 


▼ 


OWENS-CORNING FIBERGLAS 






T 


• 


• 


• 1 


»• 




* 


• 


• 


* 


• 






▼ 


PERFECT SEAL ROCK WOOL 


T 






• 




• i 


»• 


• 


* 




• 


* 


* 


• 




▼ 


▼ 


RIVERTON LIME &. STONE 


T 






• 








# 
















▼ 


▼ 


SEALTITE INSULATION 


T 






• 




• 4 


»• 


• 


* 




• 


•' 


•' 


• 








SUN VALLEY MFG. 


T 






• 






»• 


• 


• 




• 


* 


* 


• 




▼ 


▼ 


TENNESSEE PRODUCTS 


▼ 






• 


• 




»• 






• 




• 


• 






▼ 


▼ 


TEX-ARK. ROCKWOOL 


T 






• 






»• 


















▼ 


▼ 


UNITED STATES MINERAL WOOL 


T 






























▼ 


▼ 







































Figure 3-Q. Mineral Wool 



THE MARKET 

Sales of mineral wool for residential 
building insulation are dependent upon 
the volume of new residential construc- 
tion and the aggressiveness with which 
competitive materials are promoted for 
this use. In addition, mineral wool is 
used in existing dwellings and the volume 
sold for this application depends largely 
on price movements in fuels for house 
heating. Figure 3-R shows the building 
cycles over the past 30 years and the dol- 



lar volume of residential construction and 
mineral wool production (National Min- 
eral Wool Association members only) for 
the last 10 years. Mineral wool production 
and sales do not increase as fast as resi- 
dential construction. From 1944 to 1950 
there has been a 10-fold increase in both 
physical and dollar volume of residential 
building. During this time mineral wool 
production increased 3Vi fold. Conversely, 
during building downturns, insulation for 
existing houses tends to provide some de- 
gree of stability. 



204 



PRODUCT POTENTIAL 



An indication of the regions that are 
likely to experience continued growth in 
residential building can be secured from 
the relative population gains by states 
over the years. Figure 3-S shows the areas 
that have gained most in population from 
1940-1950. In general, the Pacific, South- 
western, and Mid-Central states have dis- 
played the greatest increase. Michigan's 
increase in population of 1,100,000 is 
outstanding. 

In addition to regional population 
shifts, the ability to purchase housing is 
determined by the level of income that 
exists in the area. Figure 3-T illustrates 
the regions that have made income gains 
from 1946 to 1949. The states that have 
made sizable population gains also have 
a large share of total United States in- 
come. 

The states that appear to have the best 
potential for residential building based 
on population gains and share of total in- 
come are: the Pacific Coast states, South- 
western states, Michigan, Indiana, Illinois, 
Ohio, Wisconsin, New York, New Jersey, 
Pennsylvania, and Connecticut. 

Reference to Figure 3-P, the location 
of mineral wool plants, shows that all of 
these states are served by one or more 
plants except New Mexico and Connec- 
ticut. There are fewer plants in the 
Southwestern area and on the Pacific 
Coast than population requirements would 
indicate. 

Mineral wool is a bulky material (usu- 
ally not more than 12 tons can be packed 
in a freight car) and as such, is expensive 
to ship long distances. Mineral wool can 
probably not be shipped economically 
more than 300-400 miles. Since wool can 
be produced from a great number of raw 
materials, the industry has become very 
decentralized and plants have been built 
to serve areas of dense population. The 
industry has become highly competitive, 



probably because manufacturing capacity 
may have reached the saturation point. 

Increased familiarity of the consuming 
public with the potentialities of mineral 
wool has led to a demand for better ma- 
terials to meet specific conditions. The 
manufacture of the product has tended 
to become more technical, and materials 
are produced to closer specifications. This 
fact, together with the competitive condi- 
tions, has tended to squeeze a number of 
small operators out of the business. There 
is a fairly high mortality of small plants 
when business conditions turn down. 

Changes have taken place in the in- 
dustry on the types of mineral wool that 
are most used. Figures 3-U through 3-Y 
show the production trends on loose 
wool, granulated, batts and blankets. Sales 
of batts and blankets have been increas- 
ing, loose wool has been decreasing, and 
the granulated form has not maintained 
the volume reached from 1945 to 1948. 
There is also a tendency toward insulating 
with thinner batts since FHA has ap- 
proved one inch of mineral wool as satis- 
factory for qualification under its regu- 
lations. 

From this discussion of the mineral 
wool market, the distributive function is 
the most important factor in successful 
mineral wool operation. It would be ad- 
vantageous for us to market a product in- 
volving as little fabrication as possible 
from the point of view of plant invest- 
ment and concentration of effort. Secur- 
ing distribution for loose and granulated 
wool would be difficult since demand for 
these forms has been falling off. To sell 
loose wool, we would have to interest an 
existing producer who would fabricate the 
product into batts and blankets. With 
overcapacity in the industry, producers 
would not welcome a new competitor, 
and could be interested only at a price 
below their cost of production. If we were 



PRODUCT POTENTIAL 



205 



to manufacture a complete line of min- 
eral wool products, distribution could best 
be secured by one of the companies pre- 
viously mentioned who has no plants but 
who sells mineral wool to round out a line 
of building products. While Michigan is 



ringed with a high concentration of min- 
eral wool plants, advantageously located 
to supply the more populous southern 
portion of the state, a distributor might 
find it desirable to have the output of our 
plant for sales in this area. 




1920 



1925 



1930 



1935 



1940 



1945 



1950 



SOURCE: RESIDENTIAL BUILDING- U S OEPT OF 
COMMERCE-MINERAL WOOL PRODUCTION - 
NATIONAL MINERAL WOOL ASSOCIATION 



Figure 3-R. Residential Building and Mineral Wool Production for Residential Use 




TO 10 % 
10% TO 20% 
20% TO 40% 
40% AND OVER 
Figure 3-S. Growth of United States Population, 1940-1950 (percentage of change) 




UNDER $1,000 
| ; '•.': • I $1,000 TO $1,250 
$1,250 TO $1,500 
$1,500 AND OVER 



Figure 3-T. Per Capita Income Payments to Individuals in 1949, in Dollars 



206 




UNDER 10% 
10% TO 20% 
20% AND OVER 



Figure 3-T Continued. The Percentage of Change in Per Capita Payments to Individuals, 1946-1949 




Figure 3-U. Dollar Volume of All Sales of Mineral Wool Products by 23 Companies 



207 




Figure 3-V. Mineral Wool Production of Batts and Blankets 2" or under: by Quarters 




Figure 3-W. Mineral Wool Production of Batts and Blankets 3" and over: by Quarters 



208 




Figure 3-X. Mineral Wool Production of Loose Wool: by Quarters 




Figure 3-Y. Mineral Wool Production of Granulated Wool: by Quarters 



209 



210 



PRODUCT POTENTIAL 



Prices 

Producers of loose mineral wool realize 
from $45 to $55 per ton for this product. 
Loose wool at retail brings from $70 to 
$75 per ton. Batts at the manufacturing 
level sell for $48 to $58 per 1,000 square 
feet and retail at $80 to $85 per 1,000 
square feet. An estimate of manufactur- 



ing cost per ton of loose wool would run 
around $35 to $40. 



Should Dow Chemical Company go 
ahead with plans to produce rock wool? 
In what form? How useful are the re- 
search findings in making these decisions? 
How could the research have been im- 
proved? Is any further research called for 
at this time? 



PRODUCT POTENTIAL 



211 



CASE 3-9 

Seaborg Machine Tool Company 



In March, 1955, Mr. Halloway, vice- 
president-sales, of the Seaborg Machine 
Tool Company, engaged Mr. Nathaniel 
Kidder, an industrial market research con- 
sultant, to survey the market for one of 
the company's products (machine A). 
The Seaborg company made a broad line 
of heavy machine tools which it distrib- 
uted nationally in the metalworking mar- 
ket. Seaborg did not have a marketing re- 
search department. 

Mr. Kidder believed that his first step 
should be to determine what kind of in- 
formation would prove helpful to manage- 
ment. After consulting several of the dis- 
trict sales managers he decided that they 
wanted seven types of information about 
the metalworking plants in their areas: 

Name of plant and location. 

Size of plant. 

Product manufactured in plant. 

Is plant in our market? 

Near-future purchasing plans. 

Materials being processed. 

Name and title of man specifying the 
purchase of type "A" machines. 

Mr. Kidder's next move was to analyze 
the company's past sales records from 
which he obtained the name of every 
plant that had purchased Product A in 
the last five years. In addition, he asked 
the district sales managers to check off 
those industries which had production 
processes in which they believed type "A" 
machines would be useful. 

The next step was to determine the 
"Standard Industrial Classification" of 
products made by these plants, 8 and thus 
assign S.I.C. numbers to Seaborg's actual 
and potential customers. The result was 



a list of over 100 different four-digit S.I.C. 
groups. 

This done, Mr. Kidder designed the 
questionnaire and covering letter repro- 
duced in Figure 3-Z. To select companies 
for "questionnairing," he employed the 
services of the marketing division of 
the magazine, Iron Age, which had, in 
1954, completed a census of metalwork- 
ing plants employing more than 20 pro- 
duction workers. 9 

From the S.I.C. categories designated 
by Mr. Kidder, the Iron Age marketing 
division selected and precoded 14,000 
companies. Initial returns on Seaborg 
mailings to this group were 32 per cent, 
which was raised to 65 per cent by three 
additional mailings. 

In handling these returns, Mr. Kidder's 
first step was to classify the respondents 
according to geographical area, then ac- 
cording to industry. Next the responses 
were tabulated on a blueprint with col- 
umn headings for each of the questions 
on the questionnaire (see Figure 3-AA). 



8 The Standard Industrial Classification of 
industry by products was developed for use 
in the U.S. Census of Manufactures. Code 
numbers of two, three, and four digits are 
assigned to industries and products, two digit 
numbers indicating the general industrial 
classification, four digit numbers indicating 
the specific product. For example, 20 stands 
for food and kindred products; 202 for dairy 
products; 2021 for creamery butter. 

9 Based on a continuing survey, using mail 
questionnaires and, when they are not an- 
swered, telegraph, information is assembled 
and analyzed on I.B.M. cards showing the 
company name, address, four-digit code num- 
ber, departments operated, and number of 
production workers. 



SEABORG MACHINE TOOL COMPANY 

April 11, 1955 



Castor Machine Co. 
Easton, Pa. 



To: The General Manager or Plant Manager 

Dear Sir: 

In order to assist in making better machine tools for industry, we need 
information about machines in use and the ways in which they are being 
used. It is only from companies such as yours that we can obtain this 
critical information. 

Will you, therefore, please do us a favor by passing this letter to the man 
in your organization who is best qualified to answer the questions on the 
reverse side. 

If you are not using type "A" machines or a similar machine in your 
business, please indicate that fact and return the questionnaire anyway, 
because only in that way can we be sure that our replies are representative 
of your segment of the metalworking industry. 

Many thanks for your cooperation in this matter. 

Please use the enclosed stamped and self-addressed envelope for your 
reply. 

Sincerely, 



NRK-ES 
3555-1-246 



Figure 3-Z 

1. In your business do you use type "A" machines or similar machines: 

Yes No 

IF YES: 

2. Approximately how many type "A" machines and similar machines of the following sizes 
do you use: 

Size No. of Machines 

1" 

1%" 



2" 
2%' 



3. Approximately how many hand operated type"B" machines do you use: 

4. Approximately how many automatic type "A" machines do you use: 



5. Are you planning on the purchase of new or additional type "A" machines: 

Yes No 

Hand Operated Yes No 

Automatic Yes No 

6. Approximately what percentage of your work is on the following material: 
A. Iron % B. Steel % C. Nonferrous % 

7. Your Name: 

8. Title: 

Figure 3-Z continued 



PRODUCT POTENTIAL 



213 





FIRMS WHO USE OUR 


rYPE 


OF MACHINES 










SIC. 


FIRM NAME & 
ADDRESS 


CITY STATE 
NAME & TITLE 


NO. OF 
WORKERS 


SIZE OF MACH'S USED 


TYPE "B" 
MACH'S USED 


AUTOMATIC 
MACH'S USED 


PLAN TO PURCHASE 


MATERIALS 


NO. 


1" Ik" 


2" 


m 


HAND 
OPERAIEO 


AUTO- 
MATIC 


IRON 
% 


STEEL 


NON-FERROUS 
% 


ALLE 


MTOWN-BETHLE 


HEM AREA 1 


























3522 


AGRICULTURAL h 

XYZCalnc. 
10 Main St. 


AACH. 

Al lentown 
J.Jones Plt.Mgr. 


75 


1 


3 


1 


1 


4 


1 


NO 


NO 


act/o 


59% 


1% 


3531 


CONSTRUCTION. M 

H UK Co. Inc. 
10 North St. 


NING & SIMILAR EQ 

Easton,Pa. 
J.Smith Ch.Engr. 


UIP. 

50 














NO 


NO 


(lSPECIALW&SMACH.) 


3555 


PRINTING TRADES 
ABC CO. INC. 
10 South St. 


MACH.& EQUIP. 
BETHLEHEM 
R.BROWN V.R 


246 


4 


2 


1 




10 




Yes 




60% 


20% 


20% 


3561 


PUMPS.AIR* GAS 

LMN Co. 
10 WestSt. 


COMPRESSORS 

Easton,Pa. 
S.White Supt. 


105 


1 


1 




1 


3 


1 


NO 


NO 


70% 


20% 


10% 



Figure 3-AA. Philadelphia District Survey (First Mailing) 



Next the information from the 9,120 
respondents was duplicated and forwarded 
to the district sales managers, to whom 
Mr. Kidder believed it would prove help- 
ful in several respects. In the first place 
it was likely that the manager would 
be able to determine some of the plants 
which were not in the market for type 
"A" machines, namely, those among the 
9,120 respondents answering negatively. 
At $17 to $40 per sales call, this knowl- 
edge would mean considerable savings on 
ineffective calls. In the second place, Mr. 
Kidder stated that the survey had ob- 
tained specific information on companies 
that were in the market — including data 
on the size of the plant, its principal 
product, the number and size of machines 
being used, whether they were hand- 
operated or automatic models, the kind 
of material being worked, and the name 
and title of the man involved in the de- 
cision to buy. 

Mr. Kidder's next step in handling the 



questionnaire returns was to tabulate the 
number of plants in each S.I.C. group as 
a basis for calculating what percentage in 
each group were in the market for type 
"A" machines. For example, Mr. Kidder 
stated that if 100 responses were received 
for plants in a specific S.I.C. group and 75 
per cent of them could use an "A" type 
machine, this indicated that 75 per cent 
of all plants in that group were worth 
a call. 

The final step was to build up a pic- 
ture of the total potential demand for 
type "A" machines. In this instance, Mr. 
Kidder took into account the varying size 
of plants in the market on the ground 
that he believed that within a given four- 
digit S.I.C. there was a positive relation- 
ship between the number of production 
workers in a plant and the number of 
type "A" machines used. 

In carrying out this phase of his job, 
Mr. Kidder relied on the published sum- 
mary version of the Iron Age metalwork- 



214 



PRODUCT POTENTIAL 



ing census entitled, Basic Marketing 
Data. 10 Starting, for example, in the Chi- 
cago area with S.I.C. group 3543 (cutting 
tools, gauges, dies and accessories), he 
found 80 plants listed in this category. As 
38 per cent of all plants in this group had 
been found to have use for type "A" 
machines, a figure of 30 plants was taken 
as the Chicago area potential for group 
3543. 

Similar calculations were made for all 
S.I.C. groups to obtain the total potential 
for the Chicago area. By the same process 
the potential of other areas was computed 
and the figures were totaled for the coun- 
try as a whole. 

On the basis of industrial area and na- 
tional potentials, the Seaborg company 
realigned its sales effort. In some cases, 
the sales staff was increased or distributors 
were replaced by company salesmen. In 



others, salesmen were replaced by distrib- 
utors. For example, the number of sales- 
men in the Chicago office was increased 
from four to six while salesmen in Wis- 
consin and Iowa were replaced by dis- 
tributors. 



To what uses should the findings of 
this research be put? Was the Seaborg 
company right to realign its sales effort? 
Is this a sound way to estimate national 
and/or regional potentials? 



10 Here the country is divided into industrial 
areas; and for each such area, statistics in- 
clude number of plants by S.I.C. and size 
groups, plant worker employment, number 
of manufacturing departments for each of 
24 types (stamping, welding, and so on), 
weight indexes, and so forth. 



PRODUCT POTENTIAL 



215 



CASE 3-10 

Research for a New Magazine 



This report summarizes a research 
study made in 1958, for the purpose of 
determining the marketability of a pros- 
pective weekly newsmagazine. The study 
was designed to provide some indications 
as to whether the prospective newsmaga- 
zine could be commercially feasible. 



NEWS 

The prospective newsmagazine, tenta- 
tively titled News, would be a weekly 
publication and would be generally simi- 
lar to the three established newsmaga- 
zines, Time, Newsweek, and U.S. News 6 
World Report. It would attempt to at- 
tract readers and subscribers by offering a 
high quality of writing and reporting, and, 
in particular, by emphasizing certain dif- 
ferentiating features. 

The foremost distinctive feature of 
News would be an editorial policy of un- 
biased and objective journalism. No spe- 
cial interests would be served by News, 
nor would distortion of facts be per- 
mitted. This is not to say that opinions 
would be nonexistent in News. Subjec- 
tivity would be frequently included, but 
the authors of the opinions would be 
identified through the use of by-lines. 
Thus, another distinguishing feature of 
News would be signed articles. By the use 
of signed articles, the editors of News 
would endeavor to avoid "group journal- 
ism" which could produce inaccuracy and 
distortion as well as monotony. 

The prestige of well-known journalists 
serving as staff and feature writers would 
be a favorable characteristic of News. 



High-caliber journalists could be expected 
to be attracted to News by the offer of 
by-lines as well as by good pay. 

Another distinctive characteristic of 
News would be its newness, both in ap- 
pearance and in content. A new type 
face and a cover of heavy uncalendered 
stock would be visual evidence of the 
uniqueness of News. In addition to signed 
articles, another element in News's pres- 
entation would be an editorial style gen- 
erously flavored with humor. Special serv- 
ices such as weekly news summaries and 
full tests of major documents would be 
made available to the readers of News. 

Finally News would be differentiated 
from the present three newsmagazines by 
a higher single-copy price on the cover 
and a higher subscription price. It was 
expected that News would sell for 35 
cents, 10 cents more than other news- 
magazine prices at that time. 

OBJECTIVES 

As stated previously, the general pur- 
pose of the research study was to deter- 
mine whether News, with its format as 
proposed, could be commercially feasible. 
However, the specific objectives of the 
study were to test certain propositions, as 
well as to provide some secondary infor- 
mation. 

The first proposition was that "in gen- 
eral, people object to the idea of opinion 
being included in news reporting, and will 
tolerate subjectivity only if its author is 
identified." The second proposition was 
that "most people recognize that the three 



216 PRODUCT POTENTIAL 

NEWS RESEARCH PROJECT 



We shall appreciate your frank answers to the following questions. Most of them can be answered 
by a check mark. The numbers in parentheses are there only to assist us in tabulating the results. 



1. Which of the following do you use regularly as a source of news? Please check 



one or more. 



Radio News (l-l) Sunday Newspaper (1-4) 

Television News (1-2) Weekly Newsmagazine (1-5) 

Daily Newspaper (1-3) Other (Please specify) 





(?-1) 




(?-?) 


Daily Newspaper 


(2-3) 



2. If you had to give up all but one source of news, which one would you keep? Please check one. 

Sunday Newspaper (2-4) 

Weekly Newsmagazine (2-5) 

Other (Please specify) 

(2-6) 

3. Please indicate how you use the news sources listed below by checking the appropriate box or 
boxes. 

Complete Particular All Sides 

Highlights Coverage Viewpoint of Story 

Radio News . (3^ 

Television News (4) 

Daily Newspaper (5) 

Sunday Newspaper (6) 

Weekly Newsmagazine . (7) 

Other (Please specify) 



(8) 



4. Please list your favorite news program or news publication in each category below. 

Radio News (9-1) 

Television News (9-2) 

Daily Newspaper (9-3) 

Sunday Newspaper (9-4) 

Weekly Newsmagazine (9-5) 

Other (Please specify) 



(9-6) 



5. Who is your favorite news reporter or correspondent working in each of trie following news 



Radio News . (10-1) 

Television News UO-2) 

Daily Newspaper (10-3) 

Sunday Newspaper (10-4) 

Weekly Newsmagazine (10-5) 

Other (Please specify) 

(10-6) 

Figure 3-BB 



PAGE 2 



6. By checking the appropriate box please indicate where on each of the following scales you 
would rate the three news magazines as news sources. We would appreciate your responses to 
all three magazines even though you may not be thoroughly acquainted with all of them. (The 
higher the number the more strongly the description applies to the magazine. Check the box 
under if you think neither description applies to the magazine.) 



a. U. S. News & World Report 

3 2 

Dull 



10 1 2 3 

□ □□□□□□ 



Serious 

Accurate 

Complete Coverage 

Pro-Business 

Pro-Labor 

Pro-Republican 

Objective 

b. Time magazine 
Dull 

Serious 

Accurate 

Complete Coverage 

Pro-Business 

Pro-Labor 

Pro-Republican 

Objective 

c. Newsweek magazine 
Dull 

Serious 

Accurate 

Complete Coverage 

Pro-Business 

Pro-Labor 

Pro-Republican 

Objective 



□ □□□□□□ 

□ □□□'□□□ 

a □ □ □ □ □ □ 

□ □□□□□□ 

□ □□□□□□ 
a □ □ □ □ □ □ 

□ □□□□□□ 

3 2 10 12 3 

□ □ □ □ □ □ □ 

□ □□□□□□ 

□ □□□□□□ 

□ -□ □ U2 □ □ □ 

□ □□□nee 

□ □□□□□□ 
□□□□□.□□ 

□ □□□□□.□ 

3 2 10 12 3 

□ □□□□□□ 

□ □□□□□□ 

□ □■□□□□□ 

□ □.□□□'□□ 

□ □□□□'□□ 

□ □ c □ □ □ □ 

□ U2 C C □ □ C2 

□ □□'□□□■□ 

Figure 3-BB Continued 



Interesting 

Amusing 

Unreliable 



(11) 

(12) 
(13) 



Highlights of News (14) 



Anti-Business 


(15) 


Anti-Labor 


(16) 


Pro-Democrat 


(17) 


Slanted 


(18) 


Interesting 


(19) 


Amusing 


(20) 


Unreliable 


(21) 


Highlights of News 


(22) 


Anti-Business 


(23) 


Anti-Labor 


(24) 


Pro-Democrat 


(25) 


Slanted 


(26) 


Interesting 


(27) 


Amusing 


(28) 


Unreliable 


(29) 


Highlights of News 


(30) 


Anti-Business 


(31) 


Anti-Labor 


(32) 


Pro-Democrat 


(33) 


Slanted 


(34) 



PAGE 3 

7. Which of the following types of stories in news magazines should include interpretation or per- 
sonal/editorial views in addition to the facts? Please check one or more in each category of 
story. 

Personal 

Interpretation or Editorial 

Facts Only and Analysis Viewpoint 

War Story (35) 

Revolution (36) 

Missile Launching (37) 

Scientific Discovery (38) 

Political Campaign (39) 

Labor Racketeering (40) 

New Book, Play or Movie , (41) 

Death of a Statesman (42) 

8. Should a news story in a news magazine carry the writer's name under the following conditions? 

Yes No Doesn't Matter 

Only Facts (43) 

Facts Plus Interpretation 

Facts Plus Interpretation 
Plus Personal Viewpoint 

9. Do the following news magazines express opinions in their news articles? 

Yes No Don't Know 

Newsweek < 44 > 

Time 

U. S. News & World Report 



10. Should expressions of opinion in news magazines and newspapers be confined to editorial 
pages and signed columns? 

Yes No Don't Care 



11. What changes would you like to see in news magazines, if any: 



12. How much more per copy than present prices would you be willing to pay for a news magazine 
with . . . ? 

Otf _5£ 10£ 15£ 20£ 25£ More than 25^ 

More Pictures < 46 > 

More Details < 47 > 

More Humor < 48 > 

(49) 

(50) 

(51) 



Wider Coverage 
Less Opinion 
Signed Reporting 



13. Without looking it up, approximately how much do you think you have to pay for a single year's 
subscription to a news magazine? $ 

Figure 3-BB Continued 



PRODUCT POTENTIAL 



219 



PAGE 4 

14. Please indicate if you read any of the following magazines and how you obtain them. Check 
where appropriate. 

Read Read Buy at 

Regularly Occasionally Subscribe Newsstand 



Atlantic 

Business Week 

Fortune 

Life 

Look 

Newsweek 

New Yorker 

N. Y. Times Magazine 

Time 

U. S. News & World Report 

Foreign Periodicals 

Other (Please specify) 



(52) 
(53) 
(54) 
(55) 
(56) 
(57) 
(58) 
(59) 
(60) 
(61) 
(62) 



15. Which magazine do you think the people listed below would read? 

Life Newsweek New Yorker Time 

Newspaper Reporter 

Professor 

Bus Driver 

Judge 

Business Executive 

Union Leader 



U. S. News, etc . 



(64) 
(65) 
(66) 
(67) 
(68) 
(69) 



16. Information on the following questions would be helpful to us in classifying the results of the 
questionnaire. Please check one answer in each group which best applies to you. 



a. Size of city or town 

Less than 5,000 
5,000 to 50,000 
50,000 to 100,000 
100,000 or Over 



b. Level of education Attended: 

Grade School (71) 

High School 

College 

Graduate School 



c. Age 



Less than 20 
20 to 30 
30 to 50 
50 to 65 
65 or Over 



d. Income per year 

Less than $5,000 
$5,000 to $7,500 
$7,500 to $10,000 
$10,000 to $20,000 
$20,000 or Over 



Sex 



f. Marital Status 



Male 
Female 



(74) 



g. Occupation of chief wage earner 

Professional (76) 

Manager or Executive 

Clerical or Sales Work 

Skilled 

Unskilled 

Other (Please specify) 



Married 
Single 

h. Political affiliation 



(75) 



Republican 

Democrat 

Independent 

Other (Please specify) 



Figure 3-BB Continued 



220 



PRODUCT POTENTIAL 



present newsmagazines intrude editorially 
into the news." The final proposition was 
that "many people are concerned about 
bias and distortion in the present news- 
magazines, and are sufficiently dissatisfied 
that they would seriously consider pur- 
chasing a newsmagazine which avoided 
these faults." 



METHODOLOGY 

To achieve the objectives of testing the 
propositions and providing secondary in- 
formation, a four-step plan or methodol- 
ogy was set up. The steps were personal 
interview, pretest, final questionnaire, and 
follow-up of nonrespondents by telephone 
calls. 

[In what follows, all details not refer- 
ring to the final or main questionnaire are 
omitted.] 



FINAL QUESTIONNAIRE 

The final questionnaire was mailed to 
a sample of 5,439. The entire sample con- 
sisted of two basic groups: 4,939 names 
were obtained from a mailing house, 
which was instructed to select randomly 
approximately 820 names within each of 
six postal zones with a median educational 
level of at least 12 years; and 500 names 
were taken from the latest edition of 
Who's Who (picking one name from 
every sixth page and rotating the position 
of the name on the page). 

Up to the cutoff date, the researchers 
had received a 19.6 per cent return from 
the large mailing, and a 25.0 per cent re- 
turn from the Who's Who sample. 

The questionnaire, a four-page folder, 
is reproduced in Figure 3-BB. The results, 
keyed to the question numbers, were as 
shown below. Wherever used, the word 
"No." refers to number of answers in each 
instance — sometimes, because of nonan- 



swers, less than full number of respond- 
ents; sometimes, because of multiple an- 
swers by same respondents, more. Most of 
the tabulations are based on 908 usable 
returns from the larger mailing; a few 
graphs also show contrasts with returns 
from the smaller Who's Who mailing. 



1. 



Total 



2. 



Total 



3. 



4. 



Radio 

Television News 
Daily Newspaper 
Sunday Newspaper 
Weekly Newsmagazine 

Total 



No. 



Radio News 


365 


14% 


Television News 


425 


17 


Daily Newspaper 


654 


26 


Sunday Newspaper 


498 


19 


Weekly Newsmagazine 


539 


21 


Other 


77 


3 



2,558 



No. 



908 



No. 



100% 



Radio News 


104 


12% 


Television News 


51 


6 


Daily Newspaper 
Sunday Newspaper 
Weekly Newsmagazine 
Other 


585 
18 
83 
12 


65 
2 
9 

1 


No Answer 


55 


5 



100% 



Radio News 


540 


60 ( 


Television News 


415 


46 


Daily Newspaper 


807 


90 


Sunday Newspaper 


540 


60 


Weekly Newsmagazine 


489 


54 


Other 


67 


1 


Total respondents 


908 





26% 
40 
18 
9 

7 

100% 



PRODUCT POTENTIAL 



221 







Complete 


Particular 


All Sides 








Highlights 


Coverage 


Viewpoint 


of Story 


Total 




% 


% 


% 


% 


No. 


% 


Radio News 


74% 


13% 


7% 


3% 


677 


100% 


Television News 


68 


15 


8 


9 


662 


100 


Daily Newspaper 


19 


55 


10 


16 


858 


100 


Sunday Newspaper 


20 


45 


17 


18 


660 


100 


Weekly Newsmagazine 


14 


35 


26 


25 


698 


100 


Other 


8 


20 


36 


16 


50 


100 



6. Answers are shown graphically in Figure 3-CC. 



7. 






Plus 








Plus 


Personal or 








Interpretation 


Editorial 






Facts Only 


and Analysis 


Viewpoint 


Total 




% 


% 


% 


% 


War Story 


33% 


54% 


13% 


100% 


Revolution 


26 


57 


17 


100 


Missile Launching 


49 


41 


10 


100 


Scientific Discovery 


32 


54 


14 


100 


Political Campaign 


28 


35 


37 


100 


Labor Racketeering 


32 


38 


30 


100 


New Book, Play or Movie 


13 


34 


53 


100 


Death of a Statesman 


26 


21 


53 


100 



Note: In scoring this question, which 
asked the respondent to check the treat- 
ment he or she considered a particular 
type of story should receive in a news- 
magazine, the researchers assumed that 
the three alternatives posed a logical 



gradation of editorial intrusion, and there- 
fore recognized in each line (that is, for 
each kind of story) the checkmark which 
was farthest to the right. For a graphic 
picture, see Figure 3-DD. 



8. 



Only Facts 

Facts plus Interpretation 
Facts plus Interpretation plus 
Personal Viewpoint 



Yes 

% 

27% 
72 

85 



No 

% 

30% 
7 



Dosen't 
Matter 

% 

43% 
21 

11 



Total 



100% 
100 

100 



222 



PRODUCT POTENTIAL 



Yes 

% 



No 

% 



Total Respondents 

50% 15% 

72 11 

U.S. News & World Report 49 18 



Newsweek 
Time 



Don't Know 

% 



35% 

17 

33 



Total 



100% 

100 

100 





Democrats 






Newsweek 

Time 

U.S. News & World Report 


61% 9% 
83 4 
58 11 

Independents 


30% 

13 

31 


100% 

100 

100 


Newsweek 

Time 

U.S. News & World Report 


47% 24% 
69 10 
48 20 

Republicans 


29% 

21 

32 


100% 

100 

100 


Newsweek 

Time 

U.S. News & World Report 


51% 20% 
77 6 
44 29 


29% 

17 

27 


100% 

100 

100 



10. 

Percentage 



Yes 
69% 



No 

18% 



Don't Care 

13% 



11. There was a response of 38 per cent 
(345 answers in absolute terms) to this 
question. Of these, 152 (17 per cent of the 
sample) gave an answer which in some 
way related to the kindred topics of objec- 
tivity, bias, opinion, editorial viewpoint, 
signed reporting, and so forth, in every 
case coming out "on the side of the an- 
gels." The researchers felt that it would 
be dangerous to attach too much signifi- 
cance to this result (that almost half of 
those responding were concerned with 
bias) because the whole tone and tenor of 
the second and third pages of the ques- 
tionnaire emphasized and spotlighted this 
particular issue, but it would certainly 
have been disappointing if this problem 
had been entirely ignored by the respond- 
ents. 

It is possible, however, to conclude that 



certain people, at least, were deeply con- 
cerned in view of the length of their an- 
swers, spilling out from the space avail- 
able to them over the rest of the page. 
Here are some examples: 

"There should be no doubt in any ar- 
ticle about whether the content is facts, 
interpretation or personal viewpoint. It 
is too easy to confuse one category with 
the other." 

"More news and less personal and/or 
editorial viewpoint in news stories or ar- 
ticles. Greater emphasis on fair presenta- 
tion of all sides of news on important is- 
sues. More emphasis on international 
news, interpretation and analysis. Less 
editing of news stories prepared by com- 
petent writers on the scene." 

"Strictly objective writing is most dif- 
ficult to achieve — subtle slanting by 
choice of adjectives is common — also 
choice of topics can slant a regular read- 
er's attitude to the policy of the maga- 
zine. If more or all news articles were 
signed, then the public would be more 
aware of personal or organizational bias." 



PRODUCT POTENTIAL 



223 



"Would like to see regular list of bills 
before Congress and in committee and 
progress made on them similar to that 
run by the Chicago Sun-Times on state 
legislature, more details honestly reported. 
Checking foreign news about events fol- 
lowed in U.S. papers, it is surprising how 
many pertinent details we lack and they 
have, and we have while they lack them." 

"More objectivity and documentation. 
More joint consideration of social and 
physical sciences. Less trivia. Better por- 
trayal of man's economic, sociological, 
psychiatric, scientific, political, military, 
and so forth (sic) dilemmas and achieve- 
ments. More of the genuine eclectic." 



"It seems that very few, except British 
publications, can stick strictly to facts, 
full stories, objectively told. Let their per- 
sonal ideas, opinions, suggestions appear 
separately — not mixed in with actual 
facts. All seem very conscious of offend- 
ing advertisers, customs, and religious 
groups. There is a great need for a weekly 
newsmagazine, printed along American 
lines, but with the accuracy, objectivity 
and positive approach of the Manchester 
Guardian, N.Y. Times, or Christian Sci- 
ence Monitor. The present ones are 
mostly sensational, trashy, frequently in- 
accurate, and appeal to emotions. Not up- 
lifting and positive." 



12. 

















More than 


Total 


Total respondents 


0? 


5C 


10? 


15? 


20? 


25? 


25? 


No. 


% 


More Pictures 


65% 


16% 


14% 


2% 


1% 


1% 


1% 


529 


100% 


More Details 


53 


18 


16 


4 


2 


5 


2 


525 


100 


More Humor 


66 


16 


10 


3 


2 


2 


1 


529 


100 


Wider Coverage 


43 


16 


21 


6 


4 


7 


3 


519 


100 


Less Opinion 


64 


10 


13 


4 


2 


3 


4 


512 


100 


Signed Reporting 


67 


13 


11 


3 


1 


3 


2 


509 


100 


Regular Readers of Any Newsmagazine 
















More Pictures 


66 


15 


13 


2 


1 





3 


318 


100 


More Details 


52 


19 


15 


4 


2 


5 


3 


314 


100 


More Humor 


70 


18 


6 


1 


2 


2 


1 


317 


100 


Wider Coverage 


39 


17 


24 


6 


3 


8 


3 


313 


100 


Less Opinion 


60 


12 


15 


3 


2 


4 


4 


295 


100 


Signed Reporting 


63 


14 


13 


3 


1 


2 


4 


309 


100 


Nonreaders of Any 


Newsmagazine 
















More Pictures 


63 


18 


14 


3 


1 


1 





211 


100 


More Details 


55 


16 


18 


3 


1 


5 


2 


211 


100 


More Humor 


59 


12 


16 


7 


3 


2 


1 


212 


100 


Wider Coverage 


50 


14 


15 


7 


5 


7 


2 


206 


100 


Less Opinion 


69 


9 


10 


5 


2 


2 


3 


217 


100 


Signed Reporting 


72 


11 


9 


3 


1 


3 


1 


200 


100 


Democrats 




















More Pictures 


70 


9 


18 





3 








33 


100 


More Details 


46 


14 


20 


6 


3 


8 


3 


35 


100 


More Humor 


55 


18 


15 


3 


3 


3 


3 


33 


100 


Wider Coverage 


34 


19 


19 


6 


3 


13 


6 


32 


100 


Less Opinion 


49 


15 


12 


6 





15 


3 


33 


100 


Signed Reporting 


63 


9 


9 


3 





9 


7 


32 


100 



12. Cont. 



Independents 

More Pictures 
More Details 
More Humor 
Wider Coverage 
Less Opinion 
Signed Reporting 

Republicans 

More Pictures 
More Details 
More Humor 
Wider Coverage 
Less Opinion 
Signed Reporting 



71 


18 


3 


5 


51 


23 


11 


9 


58 


18 


13 


2 


38 


17 


11 


11 


59 


12 


12 


3 


60 


16 


6 


6 


53 


26 


10 


3 


46 


21 


21 


7 


65 


24 


5 


2 


45 


16 


24 


7 


74 


11 


7 





62 


20 


10 


7 



3 








3 


2 


5 





17 





5 


3 


6 


3 








3 





2 


3 


5 


3 


2 









38 


100 


35 


100 


40 


100 


35 


100 


34 


100 


35 


100 


62 


100 


61 


100 


55 


100 


58 


100 


58 


100 


60 


100 



13. 



Total 



No. 



Less than $4 


66 


8< 


$4-$6 


354 


45 


$6.01-$7 


129 


16 


More than $7 


239 


31 



788 



100% 



14. 



Atlantic 

Business Week 

Fortune 

Life 

Look 

Newsweek 

New Yorker 

N.Y. Times Magazine 

Time 

U.S. News & World Report 

Foreign Periodicals 

Other 

Total % 
Total No. 



Read 
Regularly 

% 

1% 

6 

5 
22 

9 

9 

6 

6 
19 
11 

1 

5 

100% 
1,461 



Read 
Occasionally 

% 

6% 

9 
10 
12 
11 
14 
10 

5 
10 
11 

2 





Subscribe 

% 

2% 

6 

6 
25 
10 

7 

4 

4 
18 

9 

2 

7 



Buy at 

Newsstand 

% 

7% 

9 

7 
11 

9 
13 
10 

7 
12 
13 

2 





100% 
2,769 



100% 
1,460 



100% 
1,378 



15. 



Newspaper Reporter 

Professor 

Bus Driver 

Judge 

Business Executive 

Union Leader 



Life Newsweek New Yorker 



18% 

15 

73 

16 

13 

27 



20% 
18 
7 
20 
19 
21 



16% 
19 

5 
16 
14 

5 



U.S. News 
& World 
Time Report 



21% 

26 

11 

24 

25 

22 



25% 
22 
4 
24 
29 
25 



Total 
Answers 
No. % 



1,632 
1,522 
905 
1,512 
1,683 
1,341 



100% 

100 

100 

100 

100 

100 



224 



PRODUCT POTENTIAL 



225 



16. 

Size of City or Town 

Less than 5,000 
5,000 to 50,000 
50,000 to 100,000 
100,000 or over 

Total % 
Total No. 



Age 

Less than 20 years 
20 to 30 
30 to 50 
50 to 65 
65 or over 

Total % 
Total No. 



Sex 

Male 
Female 

Total % 
Total No. 



2% 
11 

3 
84 

100% 
813 

% 

0% 
16 
43 
32 

9 

100% 
811 



85% 
15 

100% 
841 



Political Affiliation 

Republican 

Democrat 

Independent 

Total % 
Total No. 

Level of Education 

Grade School 
High School 
College 
Graduate School 

Total % 
Total No. 

Income per Year 

Less than $5,000 
$5,000 to $7,500 
$7,500 to $10,000 
$10,000 to $20,000 
$20,000 or over 

Total % 
Total No. 



47% 

25 

28 

100% 
836 



16 
51 
29 

100% 
845 



11% 

19 

17 

32 

21 

100% 
837 



Marital Status 

Married 
Single 

Total % 
Total No. 

Occupation of Chief Wage Earner 

Professional 

Manager or Executive 

Clerical or Sales Work 

Skilled 

Unskilled 

Other 

Total % 
Total No. 



14 

100% 
828 



35% 

40 

13 

5 

1 

6 

100% 
847 



What conclusions can be drawn about 
the researchers' three propositions? In the 
light of those conclusions, what are the 
prospects for the proposed newsmaga- 
zine? 11 Should any changes be made in 
editorial plans? Should it be priced above 
competing newsmagazines? Assuming ad- 
equate resources and skill in promotion, 
what circulation can it attain? 



11 At the time of the case, the circulation of 
Time was about 2,000,000; Newsweek, 
1,000,000; U.S. News & World Report, 
900,000. 



DULL 



SERIOUS 



ACCURATE 



COMPLETE COVERAGE 



PRO-BUSINESS 



PRO-LABOR 



PRO- REPUBLICAN 



OBJECTIVE 







INTERESTING 



AMUSING 



UNRELIABLE 



HIGHLIGHTS 



ANTI-BUSINESS 



ANTI-LABOR 



PRO-DEMOCRAT 



SLANTED 



Figure 3-CC 



U.S. News 



X X X Time 

• • • Newsweek 



NOTE : Graph depicts averages 
weighted from —3 to + 3. 



MISSILE LAUNCHING 



WAR STORY 



SCIENTIFIC DISCOVERY 







PLUS 




PLUS 


PERSONAL OR 


FACTS 


INTERPRETATION 


EDITORIAL 


ONLY 


AND ANALYSIS 


VIEWPOINT 



REVOLUTION 



LABOR RACKETEERING 



POLITICAL CAMPAIGN 



DEATH OF A STATESMAN 



NEW BOOK, PLAY 




Figure 3-DD 



Large Sample 



X X X Who's Who Sample 

NOTE : Graph depicts averages weighted 
from -I to+l, then arranged by 
ranking in large sample. 



PRODUCT POTENTIAL 227 



MORE PICTURES 



MORE HUMOR 



SIGNED REPORTING 



LESS OPINION 



MORE DETAILS 



WIDER COVERAGE 



O* 



5* 



IO<t 



l 



Large Sample 



Figure 3-EE 



XXX Who's Who Sample 

NOTE : Graph depicts averages weighted 
by monetary values, then arrang- 
ed by ranking in large sample. 



MORE PICTURES 



MORE HUMOR 



SIGNED REPORTING 



LESS OPINION 



MORE DETAILS 



WIDER COVERAGE 



O* 



5* 



IO* 



/ \ 
/ \ 
/ Y 








\ \ 
\ \ 

\ \ 






\ 
\ 

\ 

v 







Reader 

• • • Non- Reader 



Figure 3-FF 



NOTE : Graph depicts averages weighted 
by monetary values, then arrang- 
ed by ranking for readers. 






228 PRODUCT POTENTIAL 



BIBLIOGRAPHY TO SECTION THREE 

Boyd, Harper W., Jr., and Ralph Westfall, Marketing Research— Text and Cases. 
Homewood, 111.: Richard D. Irwin, Inc., 1956. 

Bross, Irwin, Design for Decision (statistical decision making). New York: The 
MacMillan Company, 1953. 

Brown, Lyndon O., Marketing and Distribution Research. New York: The Ronald 
Press Company, 1955. 

Corey, E. Raymond, The Development of Markets for New Materials. Boston: 
Division of Research, Harvard Business School, 1955. 

Deming, William E., Sample Design in Business Research. New York: John Wiley 
& Sons, Inc., 1960. 

Hyman, Herbert, Survey Design and Analysis: Principles, Cases and Procedures. 
New York: The Free Press of Glencoe, 1955. 

Jahoda, Marie, et al., Research Methods in Social Relations. New York: Henry 
Holt and Company, Incorporated, 1960. 

Moroney, M. J., Facts from Figures. Baltimore, Md.: Penquin Books, Inc., 1956. 

Phelps, D. Maynard, Planning the Product. Homewood, 111.: Richard D. Irwin, 
Inc., 1947. 

Wales, Hugh C, and Robert Ferber, A Basic Bibliography on Marketing Research. 
Chicago: American Marketing Association, 1956 (revised edition in prepara- 
tion). 

Wallis, W. Allen, and Harry V. Roberts, Statistics: A New Approach. New York: 
The Free Press of Glencoe, 1956. 



SECTION FOUR 



Expanding 
Demand 



We have been using the word "product" almost exclusively as the term for 
what a marketer offers to the market. However, his offering can equally be a serv- 
ice, ranging from plumbing repairs to life insurance; and indeed it can be a 
whole "bundle" of products and services, as in the case of a retail store. In fact, 
services are becoming an increasingly important (and profitable) part of the 
economy as the standard of living rises. It is important to keep this broader mean- 
ing in mind, as we go on to discuss how to expand the demand for an offering, 
even though, for convenience' sake, we continue to use "product" most of the 
time. 

We shall look first at the general principles involved; then at the manage- 
ment considerations. 



ACTION & REACTION 

Expansion of demand depends on marketing action and market reaction. 

Strategically, there are two kinds of marketing action which will expand de- 
mand: (1) action to change the external conditions to which consumers react, 
such as the style of the product or the place it is sold; and (2) action to cause 
a subjective change in the consumers themselves so they will react differently, 
that is, use of persuasion. Or it can be, and usually is, a combination of these two, 
as, for example, when a design or model change is made in an appliance or a car, 
and then promotion is used to persuade consumers of the advantage of such a 
change. 






229 



230 EXPANDING DEMAND 

As for the particular market reaction which is the planned objective of the 
action, again looking at it in terms of broad strategy, there are three compo- 
nents to analyze: 

1. The people who will be the source of the expanded demand— (a) the 
same people buying more of the product, or (b) additional, different 
people buying the product. 

2. The appeals eliciting increased demand— (a) the same appeals, intensi- 
fied or sharpened in focus, or (b) new or widened appeals. 

3. The elements of the marketing mix that are involved in the process of 
action and reaction— (a) product characteristics, (b) physical location, 
(c) price, and (d) communicability of attractiveness. 



CHANGING THE CONDITIONS 

Let us take a look at marketing action changing the external conditions 
to which consumers react. Here are some examples (classified by marketing mix 
element, appeal, and people, in that order) : 

Product-characteristics element; intensification of same appeals; same 
people— If children like a soft drink because it is sweet, making it sweeter can 
increase their demand for that product; or if businessmen travel by airplane 
because it is fast, then making it even faster will increase their use of that 
service. 

Price element; widening of appeals; additional people— Disposable plastic 
glasses that are used for convenience on picnics will, if lowered in price 
enough, also have appeal for everyday use by a large number of housewives, 
or a store that installs a budget-payment plan will begin to attract more young 
families. 

Product-characteristics element; widening of appeals; same people— Add- 
ing more variety of packaged dessert flavors or cake mixes will result in 
greater sales among the same people who already like the product; or adding 
styles and colors of dresses will result in more sales among the same people 
shopping at a store. 

Physical-location element; same appeals; same and additional people— 
Getting hosiery stocked in supermarkets will occasion impulse purchases that 
might otherwise not be made; putting industrial paints into a mill supply firm 
will lead to sales where immediate availability is important. 

Communicability-of-attractiveness element; widening of appeals; same 
and additional people— Salt packaged in a new free-flow container will have 
added appeal for existing customers and will also attract new buyers; or a 
lathe with clean, modern streamlining will attract buyers who have pride in 
their plant appearance. 



EXPANDING DEMAND 231 

These same external conditions to which consumers react will also be help- 
ful to a marketer trying to gain a "brand image" or a "store image" that will 
induce some part of the market to prefer his brand or his store over other brands 
or stores. But in this chapter we are concerned with the fundamentals of how to 
expand demand whether it be primary (basic) demand or selective (competitive) 
demand; the difference is only one of degree, though strategically it is important, 
as we shall see in Section Five, where we discuss selective demand and competitive 
strategy specifically. 



CAUSING CHANGE IN CONSUMERS 

This same observation applies also to the other major line of action for ex- 
panding demand: causing change in consumers' minds so they will react differ- 
ently—in other words, selling, advertising, and promotion in general. Such action 
is even more important in connection with selective demand, but it only works 
there because it works in connection with primary demand; that is, because it 
has a fundamental power to lead people, or free them, to do what they are 
naturally motivated to do. 



Buying Motives 

According to anthropologists, Clyde Kluckhohn and Henry A. Murray, much 
of individuals' behavior can be explained in terms of what they do to reduce 
tensions. When tensions are the result of goals to be satisfied, then, in the normal 
person, they actually serve as a mode for pleasure. In one highly relevant passage 
they point out: 

"It is not a tensionless state, as Freud supposed, which is generally most 
satisfying to a healthy organism, but the process of reducing tension, and, 
other factors being equal, the degree of satisfaction is roughly proportional to 
the amount of tension that is reduced per unit of time. The hungrier a man 
is, the more he will enjoy his dinner; the lonelier he is, the more pleasure he 
will experience in meeting a congenial friend, and so forth. Hence, some 
people will exercise in order to "work up" an appetite, or use absence as a 
means of revivifying their affections. A tensionless state is sometimes the ideal 
of those who suffer from chronic anxiety or resentment or a frustrated sex 
drive; but, as a rule, the absence of positive need-tensions— no appetite, no 
curiosity, no desire for fellowship, no zest— is very distressing." 1 
Professor Steuart H. Britt of Northwestern University, Editor of the Journal 
of Marketing, cites the following lists of human motives, as seen by psychologists. 



1 Clyde Kluckhohn and Henry A. Murray, Personality in Nature, Society, and Culture, 
(New York, Alfred A. Knopf, 1948), p. 15; revised edition 1958. 



232 



EXPANDING DEMAND 



Visceral 


Activity 


Aesthetic 




Drives 


Drives 


Drives 


Emotions 


Hunger 


Exercise 


Color 


Fear 


Thirst 


Rest 


Tone 


Rage 


Air-getting 


Perseveration 


Specific qualities of 


Disgust 


Temperature 


Rhythm 


taste, smell, and 


Shame 


regulation 


Novelty 


touch 


etc. 


Sexual 


Exploration 


Rhythm 




etc. 


etc. 


etc. 





Even more importantly, Britt cautions that such listings are only abstractions 
from the infinitely complex patterns of life's activities. 2 Man is more than psy- 
chological; he is an intelligent, and, as Aristotle observed more than 2000 years 
ago, a social animal; he has complicated himself as well as the world around him. 

Many new concepts and theories have been developed in this field in recent 
years. One of the most relevant to marketers is that of Professor A. H. Maslow of 
Brandeis University. He sets forth a hierarchy of basic needs: 

Physiological needs 

Safety needs 

Belongingness and love needs 

Esteem needs 

The need for self-actualization 

He observes that, in general, the physiological needs must first be satisfied; only 
then can safety needs come into play; and so on down the list. But he too cautions: 

"These needs must be understood not to be exclusive or single determiners 
of certain kinds of behavior. An example may be found in any behavior that 
seems to be physiologically motivated, such as eating, sexual play, or the like. 
The clinical psychologists have long since found that any behavior may be a 
channel through which flow various impulses. Or to say it in another way, 
most behavior is multimotivated. Within the sphere of motivational deter- 
minants any behavior tends to be determined by several or all of the basic 
needs simultaneously rather than by only one of them. The latter would be 
more an exception than the former. Eating may be partially for the sake of 
filling the stomach, and partially for the sake of comfort and amelioration of 
other needs. One may make love not only for pure sexual release, but also 
to convince oneself of one's masculinity, or to make a conquest, to feel power- 
ful, to win more basic affection. . . . This contrasts sharply with the more 
naive brand of trait psychology in which one trait or one motive accounts for 
a certain kind of act, i.e., an aggressive act is traced solely to a trait of ag- 
gressiveness." 3 



2 Steuart H. Britt, "The Strategy of Consumer Motivation/' Journal of Marketing, April 
1950, p. 667. 

3 A. H. Maslow, Motivation and Personality, New York, Harper & Brothers, 1954, p. 102. 



EXPANDING DEMAND 



233 



Or, according to Kluckhohn and Murray: 

"One of the many incongruent representations of reality which have re- 
sulted from our fixation on the old stimulus-response formula has been the 
notion of a personality as a more or less inert aggregate of response patterns, 
which requires a stimulus (often an external stimulus from the experimenter) 
to start it going, instead of conceiving of a matrix of incessant functional 
processes (the katabolic phase which spontaneously succeeds the anabolic 

*800 BILLION U.S. PRODUCTION BY 1971 
ADDS $181 BILLION TO CONSUMER NEEDS 



OPPORTUNITY FOR 55% HIGHER LIVING STANDARDS 



$800.0* 



(All in June I960 Prices 




GROSS NATIONAL PRODUCT $233.4 
(In BILLIONS of Dollars) 



GOVERNMENT 



( Federal S. local, including Defense) 



PRIVATE INVESTMENT |$364 



STANDARD OF LIVING 

(Total Personal Consumption) 



$38.3 



1^158.7 

1 



$170.0 



$120.0.: 



/ 55% 

/ Increase 

in Consumption 

needed to match 

'productive abilityl 



* 510.0 Total 

Personal 
Consumption 




940 



197 



DISPOSABLE INCOME 

(BILLIONS) 



$167.9 $254.0 



$354.3 



$555.0 



POPULATION (THOUSANDS) 



132,122 



151,683 



180,000 



220,000 






Based on Probable Population Growth and a minimum increase in production per capita 

Figure 4-A 



234 EXPANDING DEMAND 

phase of sleep), a brain seething with fantasies, programs, and projects, with 

tentative goals, expectations, and hopes, with fears and dreads, which lead on 

to actions which select certain regions and constituents of the environment 

and which apprehensively avoid others." 4 

In view of this complexity, it is much more helpful for marketing managers to 
think in terms of the attitudes that people have toward products or ideas. These 
represent, in individuals, the sum of the past effects of all the forces that influence 
them— physical, economic, personal, social— and are the mediating factors through 
which selling has to work, as it sets goals for people to satisfy through buying. 

The point remains that our motives do stem back to natural, innate drives; 
and we should be proud of the way our system of free buyers and free sellers has 
refined them, at the same time that we recognize the price we pay in also being 
free, sometimes, to want and to be persuaded to want things that are not worthy 
of us. Fortunately, for every example of foolishness and exploitation, there are 
hundreds of examples of good results in the form of broader interests and better 
tastes that have come about through competitive selling, from do-it-yourself work- 
shops to stereophonic music. The bill of fare provided by commercially sponsored 
television, which educators criticize (as they should) because it falls far short 
of their ideal for this powerful medium, still represents a substantial raising of the 
intellectual and cultural level of use by leisure time for the vast majority of 
the public. 

Once people feel that they can buy a product in the confidence that it will 
serve the purpose for which they pay out the money, and that they are not being 
misled or forced into buying, then making that purchase becomes one part of all 
that they do to raise their standard of living— and this is what has happened to the 
whole of society over time, in response to the action of marketers. The change in 
people's ability to live better is pictured graphically in Figure 4-A; this is a chart 
prepared by Arno Johnson, Vice President and Senior Economist of J. Walter 
Thompson Company (one of the U.S.'s largest advertising agencies) to show 
foreign businessmen not only what has been accomplished in this country through 
marketing, but also the opportunities that still lie ahead. 



Arena for Action 

Now, let us concentrate on what has become the major thrust of marketers 
in this country— causing change in consumers. 

First of all, it should be made clear that what selling does is to take attitudes 
that already are somewhat disposed toward a product or idea, and to reinforce and 
activate them. (This has a bearing on the optimum number and mix of appeals 
and brands, directed to different groups of people with different predispositions, as 



4 Kluckhohn and Murray, op. cit., p. 17. 



EXPANDING DEMAND 235 

described in Section Five.) Selling can't really change attitudes directly, (a) unless 
they are unimportant to the individual, in which case somebody else can change 
them easily too and take the business away from you again; or (b) unless people 
haven't yet had a chance to develop their attitudes, as in the case of a radically new 
product (though for more conventional new products people will be able to transfer 
the attitudes they already have for similar products); or (c) unless the mediating 
factors that cause the attitudes are themselves changing, as in the case of social in- 
fluences. 5 

Further, you cannot, except in unusual circumstances (like the fervor of war), 
rule people through their emotions alone. People must be given a reason for buy- 
ing which in their eyes is rational, no matter how importantly their emotions may 
be involved at the same time. 

Since selling must work through influence and communications, the marketer 
trying to follow a scientific approach can find much that will be helpful in the 
theoretical framework being developed by researchers in the area of public opinion 
and mass communications. 6 Thus, there is considerable evidence against the popu- 
lar notion that through knowledge of people's motivations you can somehow ma- 
nipulate them, or even trick them into buying against their will or without their 
realizing they are being sold. This is virtually impossible, according to Raymond A. 
Bauer, one of the country's leading social psychologists, because people have built-in 
powers of resistance. 7 If we think of selling and advertising as a process of com- 
munication, people can resist persuasion at four levels: exposure, perception, re- 
tention, and decision. More specifically: 

People who are not interested in an idea or product are less likely to 
expose themselves to information or messages about it. Thus, only 32% of a 
sample of male smokers were consistent readers of articles on health (including 
articles dealing with the relationship between smoking and lung cancer), 
whereas 60% of nonsmoking males read such articles. 8 Or, in marketing terms, 
the girl who has no expectation of a fur coat is less likely to see fur coat ads 
than a girl who is contemplating making such a purchase or receiving such a 
gift (or, for that matter, who has just become the owner of one, in which 
case she is looking for reassurance about her new possession). 

People who have a strong attitude or belief about an idea or product are 



5 See Joseph T. Klapper, "What We Know About the Effects of Mass Communications: 
The Brink of Hope," Public Opinion Quarterly, Volume 21, 1957-1958, pp. 453-474. 

6 See Donald F. Cox, "Keeping Informed: Clues for Advertising Strategists: II, Harvard 
Business Review, November-December 1961, p. 160; this not only covers most of the selected 
mass communications research material cited in this chapter, but goes much further. 

7 See Raymond A. Bauer, "Limits of Persuasion," Harvard Business Review, September- 
October 1958, p. 105. 

8 "The Impact of Health News on Attitudes and Behavior," Journalism Quarterly, Volume 
33, 1956, pp. 315-323. 



236 EXPANDING DEMAND 

less likely to perceive correctly a piece of information or a message which is 
not in conformance with their own. Thus, in a study of rumor, a picture in 
which a Red Cross truck was shown loaded with explosives was ordinarily 
perceived by subjects as a Red Cross truck carrying medical supplies (be- 
cause that is the way it "ought" to be). 9 Or, in marketing terms, the proud 
owner of an expensive camera is likely to distort the claims of a cheaper model 
into evidences of lack of quality (for example, "simplicity of operation" is 
distorted into "poorly focused pictures"). 

People who disapprove of or disagree with a piece of information or a 
message are less likely to retain, that is, remember, it. In one experiment, 
procommunist material was better learned and better remembered by pro- 
communists than by anticommunists; and the reverse was true for anticom- 
munist material. 10 Or, in marketing terms, the people who have had bad expe- 
rience with Product A or have strong loyalty to competing Product B, are less 
likely to remember advertisements for Product A. 

People whose attitudes or beliefs about an idea or product are still far 
from those expressed in a piece of information or a message are less likely to 
decide to accept or buy it. Thus, in another experiment, those whose attitudes 
strongly favored Prohibition were predisposed not to be influenced by argu- 
ments against Prohibition, and vice versa. 11 Or, in marketing terms, only those 
whose ideas are already close to believing that "this store carries a larger vari- 
ety of quality merchandise than any other store in town" are likely to decide 
to give it most of their patronage. 

These ways of resisting, or avoiding, persuasion are important because they 
represent the exact opposites of what selling and advertising are trying to do: build 
interest, change attitudes, secure approval, and bring about a buying decision. On 
the other hand, they have in them the seeds of action too. You can construct a 
strategy that is built, not on the people less likely to be exposed, to perceive, to re- 
tain, and to decide, but on their reciprocals: those who are more likely to have the 
predispositions that are favorable to your product or brand, which you, as a mar- 
keter, can reinforce or activate. 

In terms of behavioral theory— and practical marketing as well— people's pre- 
dispositions stem from the goals which they already have as individuals and mem- 
bers of society, and which they can be stimulated to feel even more strongly. In 
this sense, the job of the marketer is to serve goal-directed people by providing 
products in a context where the stronger the want that can be developed for a 
given item, the more satisfying its purchase will be. 



9 "The Basic Psychology of Rumor," Transactions of the New York Academy of Sciences, 
Series II, Volume 8, 1945, pp. 61-68. 

10 "The Learning and Forgetting of Controversial Material," Journal of Abnormal and Social 
Psychology, Volume 38, 1943, pp. 507-517. 

11 "Assimilation and Contrast Effects in Reactions and Attitude Change," Journal of Ab- 
normal and Social Psychology, Volume 55, 1957, pp. 244-252. 



EXPANDING DEMAND 237 



Taking the Initiative 

In sum, people are a match for any mere psychologist— mere, that is, if he 
doesn't go further than knowing consumer motivations. Just knowing consumer 
motivations does not mean being able to shape them to specific ends, such as get- 
ting people to buy more or to reach for this product in preference to that one. 
No, the psychologist must also be a strategist, or, perhaps, be even more of a 
strategist than a psychologist, because, no matter how helpful psychological in- 
formation may be, the action called for is strategic. Consider these facts: 

( 1 ) People need, want, and, for the most part, like to buy things; that is 
why there can be such a thing as demand in the first place. Here are resources 
of a positive nature. But there also is a negative side. There are counterbal- 
ancing factors that cause people to have resistance or inertia (including hesita- 
tion based on fear of being oversold and also on the desire to rely on what 
they know they can have confidence in ) . So the essence of selling is to take 
action that will build up the strength of the positive factors and /or dissipate 
the strength of the negative factors, to the point where the positive side out- 
weighs the negative side. 

(2) You cannot actually change people's minds— at least not without 
prohibitive expense and consumption of time. Accordingly, the only practical 
way to cause change is to array your forces so that people themselves decide 
to do what you want them to. And to get people to want to put their money 
in your product requires that there be some advantage in it for them (big or 
little, logical or psychological); it has to have the potential of moving them 
closer to satisfying some want. In consequence marketers continually try to 
add more value or benefit, as they think people will see it. So, despite some 
slippage and some waste, the end effect is improved products. 

Low-pressure selling owes its effectiveness to the fact that by approaching 
people in terms of some problem or need they have, you get a chance to 
demonstrate advantageously what your product can do to help them. So you 
lead, you guide, but you leave people free to make their own decision— in 
your favor, if you are successful. 

(3) Different products (or different brand features, or different ideas for 
that matter) have different reaction factors in people's minds. The potentiali- 
ties for securing acceptance, as well as the difficulties of overcoming resistance, 
vary over a wide range. So it is useful to have an understanding of consumers' 
motives. You have to know the strength of the forces you are contending with, 
and how they are deployed so you can best deploy your forces. 

But from that point on you must construct your strategy, not to counter, 
or change, people's motives, but to give them the help they need to change 
themselves: a carefully calculated combination of emotional and rational ap- 
peals so that no matter what the nature of resistance in the particular situa- 



238 EXPANDING DEMAND 

tion, it will tend to recede or disappear entirely in the light of new goals— that 
is, new or greater wants, with new or greater satisfactions. 

(4) Finally, since the seller is taking the initiative in drawing up his 
strategy, he can choose the area that he thinks will be most to his advantage. 
(For the most part, resistance is not planned or purposeful, although organi- 
zations like Consumers' Research and Consumers' Union are helping people 
to buy more deliberately. ) Here we come again to the marketing mix elements; 
for the seller has the choice of (a) making the product characteristics mean 
more to more consumers; (b) building the psychological value of the product 
so the dollar price seems relatively less; (c) drawing people to the physical 
locations where the product is on sale (and display), or getting the various 
channels and outlets themselves to give it more push; or (d) improving or 
enlarging the communication of its attractiveness. 

Or it can be some combination of these. For example, a toothpaste manu- 
facturer can try to give his "Ingredient X" more significance by getting the 
endorsement of some professional group, at the same time putting more money 
into his advertising program to consumers to get a wider hearing, and seeking 
a still further, subsidiary purpose, by enlisting the support of dentists ('Visit 
your dentist regularly"). However, as pointed out earlier, one of the four de- 
mand factors is likely to be relied on for the main thrust, thus resulting in a 
sharper focus both of the seller's effort and the market's response. 



Example: California Fruit Growers 

A classic example of expanding demand, which includes almost every possible 
variation on all four of the marketing mix elements, and changes both in condi- 
tions and in consumers, is that of the California Fruit Growers Exchange. Whereas 
oranges used to be Christmas-season luxuries for most of the country, they became 
a staple, and, in the form of juice, almost a necessity for American breakfasts. 
The CFGE effort included: 
Careful grading of fruit. 
Branding— "Sunkist." 

Improved freight car refrigeration and transit. 
Distribution of juice extractors. 
Jobber service program. 
Work with home economics teachers. 
Advertising on radio and in magazines, based on such appeals as health, 

taste, nutrition, economy, personal appearance, and success in life. 
Streetcar cards and billboards. 
Counter displays. 
Nutritional research. 
And, as happens so often in promotion and advertising, the activity of the 



EXPANDING DEMAND 239 

CFGE both benefited from a developing trend and helped to develop it faster 
and farther— decline in heavy physical labor, urbanization and suburbanization, 
mass transportation, heated homes, light breakfasts, older age distribution, hurry 
and complexity of life, interest in nutrition, and so on. It is a truism that it is 
difficult to reverse a downward trend (but not impossible if you mount a massive 
enough effort, as, for example, the cigar manufacturers have done, including mak- 
ing a drastic change in their product). On the other hand, promotional efforts 
that ride a rising trend are likely to pay off handsomely. Just the mathematics of 
the upward sales curve ensures an increasing rate of return for a given amount of 
expenditure; but on top of that, because of people's growing interest, the message 
gets through to them better, and the curve itself is reinforced in the process. 

Over the 40 years from 1907 to 1947, for example, California orange shipments 
increased from 10,000 cars to 100,000 cars, and dollar sales from $10,000,000 to 
almost $200,000,000. Thus, the California orange growers increased physical sales 
10 times, and dollar sales 20 times. And, in the process, they helped to expand 
demand even more widely, for, indirectly, their promotional effort stimulated 
Florida orange sales and sales of other kinds of fruit and juices generally. It should 
not be forgotten, either, that this whole expansion of the orange industry has, in- 
deed, been of direct benefit to the public in the sense of creating a better under- 
standing of nutrition and the importance of eating habits. 



Rational Buying 

Let us consider some of the aspects of promotional strategy in more detail. 
At this point, no distinction between personal selling and advertising will be made; 
although both have special problems of their own, the element of persuasion in 
each is basically the same. 

There is reason to think that most people are not quite the unthinking, non- 
economic, erratic, unconscious, or even sex-dominated and anxiety-ridden charac- 
ters that they are sometimes pictured to be. Let us start with the premise that a part 
of living is buying, including buying for psychological satisfactions (like a sense of 
well-being or aesthetic pleasure). Indeed, it is normal and healthy for people to 
like to buy, just as it is abnormal and unhealthy for them to dislike to buy (like 
misers and misanthropes). It follows that in selling we can— and should— deal 
with people as if we expected them to be normal and healthy rather than in need 
of mental hygiene— or at least predominantly on that side of the balance, so they 
have it in their own power to resolve any emotional conflict over buying. 

It hardly needs to be demonstrated that people like to buy. There is a sheer 
enjoyment in acquiring things, which goes far back to primitive roots. The act of 
buying is an expression of power, of mastery. Particularly when we are dealing with 
needs and wants above the physiological level, moving toward a goal can be as 
pleasurable as actual satisfaction. Also, people generally like to be nice to people; 



240 EXPANDING DEMAND 

since a salesman is a person, the inclination is to be nice to him; and the way to 
be nice to a salesman is to buy from him. 

Further, the person who is normal and healthy-minded tends to do what so- 
ciety approves of. The purchase of goods and services— not just food and clothing 
but education, and not just a car but a car of modern fashion— is part and parcel 
of our way of life. And it is not only socially right to buy certain kinds of individual 
products and services; it also is socially right to buy lots of things in general. In- 
deed, this is how people's success is usually measured— as individuals or as families, 
in their own self-opinion or in the eyes of society. Or, at least, this is how most 
people feel, which is what counts. 

There is, then, a positive urge to buy— and a negative counterforce. Just as it 
is personally and socially acceptable to buy, so it is not socially acceptable to over- 
buy, that is, to the point where a person appears reckless, improvident, a poor 
manager of his finances; or to buy unwisely, that is, without due regard for price 
and quality. Admittedly, these are vague terms; in fact, they shift with time, and 
at any one time mean different things to different people. But the idea that there 
is a point beyond which it is not right to buy, is, nonetheless, very real in every indi- 
vidual's mind, and people are uneasy lest they go beyond it without realizing they 
are doing so. Hence, they are wary of pressure or trickery or of their own bad 
judgment. 

In other words, people must have reassurance that they are making a good 
buy. This need is accentuated because people never can be completely satisfied 
with a purchase. When a person chooses between two or more alternatives (it 
could be alternative ideas or products, or simply whether to buy or not buy), he 
knows that there are certain advantages on the side of the decision he has made; 
that is why he made that decision. But he also knows there are some advantages, 
though presumably lesser, on the side of the alternative he did not choose (par- 
ticularly if he has been subject to advertising or promotion in its behalf)— which of 
course means some disadvantages for the one he did choose. 

This is allied to the theory of "cognitive dissonance" developed by Professor 
Leon Festinger of Stanford University. According to his theory, the individual, 
faced with "dissonance" between his cognition of the advantages or appeals 
in behalf of one product or idea and his cognition of his own preference or 
decision in favor of another product or idea, has to take some action to reduce the 
tension— ranging from refusal to hear or see the message to acceptance of the 
new product or new idea. 12 Needless to say, it is the last-named action that sellers 
seek; but, once again, they have to give the individual some strong reassurance 
of the wisdom or propriety of such action. 

In general, every buying motive is emotional to some extent, since it is 
related to attitudes and predispositions; and if tension is involved, it is doubly so. 



12 Leon Festinger, A Theory of Cognitive Dissonance (Evanston, Illinois, Row, Peterson 
and Company, 1957). 



EXPANDING DEMAND 241 

By the same token, to the extent that a person is making a buying decision that 
involves emotional strain (as would always be somewhat so when the decision 
represents a change from existing products or shopping habits in response to a 
selling attempt), to that extent the way to confirm a person's buying decision al- 
most has to be rational, because only if it is rational will it be believable and 
acceptable as a way of reassuring the person in that emotional situation. In this con- 
nection, it is significant that rational arguments recognizing both sides of a ques- 
tion have been found to be much more effective than one-sided arguments with 
people who were emotionally opposed to an idea, while more emotional one-sided 
appeals worked quite satisfactorily with those who were already emotionally in 
favor. 13 

We use the term rational here from the viewpoint of the buyer, not from the 
viewpoint of what the marketer or anyone sitting in judgment thinks is rational. 
It is what people feel and think that determines how they react to the market, not 
what the marketer feels and thinks. The person who decides to do something that 
will get him closer to his goal is behaving rationally. (Whether his goal is a "good" 
one or not is another question.) Snygg and Combs even go so far as to say that "the 
drunk who turns on the light so that he can hear better is, from his point of view, 
acting rationally." 14 Fortunately, people are usually not that fuzzy or irresponsible, 
and (except for the minority that have mental problems) their rationality is that 
of thinking beings with a sense of dignity for themselves and of deference for the 
society in which they live. In such a context, it may be helpful to keep this defi- 
nition in mind: 

Rational applies to reasons for buying (or not buying) which are self- 
approved— that is, which the buyer feels to be right and reasonable because 
they are in line with his own expectations of himself as a thinking man and /or 
his understanding of what other people (society) would consider to be right 
and reasonable on the part of a thinking man. 

Emotional is not an exact contrast to rational, since the very fact that people 
feel the need to have some reason for buying reflects an emotional drive. But 
emotional, as applied to buying motives or selling appeals, connotes the direct 
satisfaction of wants or desires without regard for whether they may be appropriate 
for a thinking man. 

One other point needs clarification. Very often a distinction is made between 
personal forces and social forces as determinants of behavior. In a very real sense, 
however, the two are closely related: one big reason for a person's being concerned 
to do what he feels society expects is that he wants to "actualize" himself (or 



13 Carl I. Hovland, Arthur A. Lumsdaine, and Fred D. Sheffield, Experiments on Mass Com- 
munications (Princeton, Princeton University Press, 1949). 

14 Donald Snygg and Arthur W. Combs, Individual Behavior (New York, Harper & Brothers, 
1949), p. 13. 



242 EXPANDING DEMAND 

realize his full potential) as a member of society. Not that a buyer thinks this 
all out, or even formulates it explicitly, but, as a result of all his experiences and 
the current norms of society, he just feels some reasons for buying are "all right" 
for him, and others are not. 

There is a similar requirement for a marketer to offer his product at a place, 
in a form, and with a price that people feel to be right (that is, in line 
with their expectations, or at least not radically out of line). But the marketer's 
problem— or opportunity— is even greater in connection with the reason for buying 
that people must have. The reason is all the stronger if it is right, in the sense of 
being in line with individuals' goals, including the goal of being an accepted mem- 
ber of society. 

Thus, we can construct a strategy to minimize resistance and maximize the 
urge to buy, by planning and presenting rational goals for people which (a) will 
lead them to the particular product or brand, in such a way that they satisfy their 
motivations, and even act as their own psychiatrists in the event of any conflict, 
and (b) will introduce or intensify the kinds of emotional drives that push or pull 
in the same direction, when that is appropriate too. The rational part is indispensa- 
ble; the emotional part is valuable when it applies, but it does not fit in all cases. 

Take the rational aspect, first. For every selling situation the buyer must have 
some self-approved reason for saying yes— whether the affirmation is in terms of 
handing over the money or signing on the dotted line, or just feeling agreement 
with an advertising message. For example, a man may want to buy a new car 
because it is new and sleek and shiny, but he also needs some rationalization like, 
"I'll get a better trade-in if I buy now." Or a woman may want a particular brand 
of soap because she thinks it will make her seductive to all mankind, but she also 
must be able to think something like, "If it makes me look prettier, my husband 
will be pleased"; or more simply, "It will protect my skin"; or still more simply, 
"It cleans." 

This is why it sometimes is difficult for a salesman to "close" a sale, even 
though the prospect may seem strongly attracted to the product; a convincing 
rational reason has not been offered, even though the actual motivation for buying 
may have been amply satisfied. The same thing is true in direct mail selling; a man 
may want to subscribe to a magazine because it makes him feel important, but 
the promotional piece must also assure him that by filling out and mailing the 
subscription order form he will receive some practical help in improving his pro- 
fessional skills or solving his business problems. 

This is also true in advertising. Of two ads, both of which apply to a product 
mainly bought for emotional reasons, the one that provides in addition some co- 
gent rational reason usually produces a stronger reaction than the one that does 
not. One explanation of why the combination of a strongly emotional advertising 
appeal to the dream of masculinity and a new crush-proof box worked so well for 
Marlboro cigarettes is that the box offered a generally acceptable rational reason. 



EXPANDING DEMAND 243 

In a very real sense, then, the need for a rational goal is overriding; that is, it 
applies no matter how emotional, noneconomic, or unconscious the motivations 
are in the particular buying situation. 

True, rational reasons are more likely to be conscious than reasons which are 
not self-approved, because there is no cause for shame or anxiety about them and 
hence no tendency to push them into the unconscious. This is fortunate, because 
it means that rational motives (as defined here) are that much easier to identify 
for selling strategy. However, the fact remains that rational motives can be uncon- 
scious, particularly when conflicting motives toward a purchase are involved, as in 
this example: 

A man may want to buy a new labor-saving device for his wife, but 
may be very concerned about the money outlay. Prudence in spending money 
is a rational motive, but the man also likes to think of himself as being good 
to his wife. So what does he do? He finds fault with the machinery of the 
proposed new washing machine (or whatever it is)— and does not admit to 
his wife or to himself that a technical defect is not the real reason, and he is 
simply holding back because he doesn't feel right about spending the money. 

Even more important, a rational motive may apply also to a noneconomic or 
intangible value so long as it is self-approved. Most people would consider it 
reasonable that a man should buy helpful things for his wife, as in the example 
just cited; the fact that he and society think that it is one of his responsibilities 
makes it reasonable— even though it may be a self-pleasure, too. Similarly, in this 
sense, it is rational to want to live in a good neighborhood, to have a modern-style 
car, and to own a television set. 

One of the great buying phenomena of modern times is the way it became 
almost obligatory to buy a television set so the children of a family could hold 
their own in their relationships with other children— an intangible but very ra- 
tional reason. It may well be that people already strongly desired TV sets for un- 
approved, selfish reasons, and it was the overwhelming availability of a good ra- 
tional reason for justifying the purchase which triggered the buying wave. 

It is interesting to note that when one large manufacturer mounted an exten- 
sive newspaper campaign playing on the theme of ''the loneliest kid in the block" 
—the child without a television set in his home— public indignation forced 
cancellation of the campaign. The theme was so powerful that it put families 
who could not afford a TV set at that time to shame. But the fact that the manu- 
facturer showed unpardonably poor taste in going to such an exreme in this in- 
stance does not alter the soundness of the basic method. Note how much less 
rapidly color TV has made its way; circumstances do not permit the same strik- 
ing rational appeal of doing what is good for one's children. (Also dealers were 
initially unenthusiastic about the product, in part because they were not expe- 
rienced in coping with its greater technical complexity.) 



244 EXPANDING DEMAND 



SELLING STRATEGY 



One of the advantages of a concept which makes a clear-cut distinction be- 
tween what is self-approved and what is not self-approved (whether it is called 
rationality or something else) is the way it cuts through a lot of confusion and 
pinpoints the crucial factor in selling strategy. 

To illustrate: 

Take a man (prior to the swing to compact cars) who honestly wants to 
buy a foreign car for the sake of economy, but is afraid he will look like an 
oddball. So he pretends, to the rest of the world and to himself, that he is 
thinking of how easy it will be for his wife to park. Being kind to the weaker 
sex is his only course; he cannot follow the economic motive because it is not 
self-approved. At this time and in this neighborhood, there is something wrong 
with a man who doesn't have a big car like everybody else's. 

But what is socially acceptable is always changing, and we will do well 
to keep this in mind as market researchers report "changing" consumer pref- 
erences. We are justified in wondering if some of the "changes" recently re- 
ported may not simply be due to differences in the freedom with which con- 
sumers feel they can talk about specific products and advertisements. Thus, it 
now is popular to boast about the economies offered by small cars. 

This does not mean that economy has become a more important motive 
than it used to be. It only means that economy has become a socially ac- 
cepted and therefore self-approved reason for buying a car. 

We must recognize that people often advance rational reasons for actions that 
actually reflect other motives. Or, to put it another way, rational reasons often 
cover other motives. To illustrate: 

It is rational for a woman to patronize a supermarket or a discount 
house for the sake of economy, but this is not altogether her real motive. Cer- 
tainly she is also influenced by her desire to appear (to herself and to the 
world) as a careful buyer or a shrewd bargainer— perhaps even more so than 
by her desire for actual money savings. (The lure of trading stamps, worth 
one-tenth of a cent each, would seem to be in line with this interpretation.) 
And if sjie does not patronize such outlets, she can always find equally ra- 
tional reasons— "The meats are so much better where I shop"; or, "They 
stand back of their products at my store." 

There is no doubt that, no matter how rational people try to be, very often 
they buy (or do not buy) things for emotional reasons. But it does not follow that 
the best way to sell them is by appeals to those emotional motives. The very fact 
that people insist on devising some rational explanation for themselves suggests 
it may be effective to supply them with a strong rational reason, so they can 



EXPANDING DEMAND 245 

follow their emotional bent to the seller's product more surely and speedily. 
On the other hand, it may actually be ineffective to use emotional appeals because 
they may have more difficulty in securing positive attention, rather than just mo- 
mentary, vague reaction or, if they do get across, may only serve to point out 
motives the buyer cannot be proud of as the thinking man he wants to be. 

There is much discussion these days about consumers' self-images and product 
images (as well as store and brand images). The implication is that if the product 
image is in line with a consumer's image, a buying urge will result. Certainly, 
here is a useful way of describing relationships between product and buyer, but 
not necessarily of creating such relationships. For example, 65 per cent of the 
people who own a particular make of automobile may think of it as the sporty, 
powerful kind that commandoes and test pilots would drive; this may be their 
way of signifying that something they have already chosen (for whatever reason) 
is in line with their own needs or dreams, but it doesn't follow that advertising it 
this way (say, with rakish drivers and scenes of derring-do) will necessarily cause 
other people to develop the same feelings of wanting it and therefore deciding to 
buy it. 

Is a product image the cause of new buying, or the effect of good selling in 
the past? Tests indicate that people usually show a strong preference for the prod- 
ucts they have already chosen. 15 Is a self-image a choherent unit, or just a bundle 
of unsorted and unweighable motivations? No matter what the answer to these 
questions, a large part of the job of making the product image fit the consumer's 
self-image can be done, and may have to be done, at the rational level. After all 
one of the biggest components of a person's self-image is the picture of himself 
as a reasonable, thinking man. 

This may seem like doing injustice to the scope and subtlety of psychological 
analysis, but a certain amount of such oversimplification is basic to the construc- 
tion of efficient selling strategy. For, as we have seen, no matter what else is done 
or what other motivations are involved, to close a sale or make a telling adver- 
tising impression, the buyer must be given a self-approved reason for purchasing 
or preferring the product. 

People's minds are complex, and every individual's self-image is different from 
every other individual's self-image. As Kluckhohn and Murray say, every man is in 
certain respects (a) like all other men, (b) like some other men, and (c) like no 
other man. 16 But selling must be a concentrated, focused action. So the seller must 
concentrate on the strongest (or most effectively communicable) buying motive 
(or set of buying motives) for the particular consumers he plans to sell to, and 
count on it, whether there are other reasons for buying or for not buying, to max- 
imize the buying urge and/or minimize the resistance. (However, there will be 



15 Jack W. Brehm, "Post-Decision Changes in the Desirability of Alternatives/' Journal of 
Abnormal and Social Psychology, Volume 52, 1956, pp. 384-389. 

16 Kluckhohn and Murray, op. cit., p. 35. 



246 EXPANDING DEMAND 

times when those with different motives offer a large enough market to justify a 
different set of appeals.) 

In view of these circumstances, a rational motive is likely to be the most 
strategic motive to use, or at least an indispensable part of the most strategic 
set of motives to use. Thus: 

A. Suppose there is a rational buying motive in your favor and no real 
resistance beyond the routine desire to be sure to make a good buy— as, for ex- 
ample, when a man honestly wants a car that will provide the most eco- 
nomical transportation to work. Here you can demonstrate to him that your 
car will provide that kind of transportation, reinforcing his existing rational 
motive; and perhaps add some emotional warmth to give it more intensity. 

B. But suppose the prospect's buying urge is emotional. Professor Bauer 
cites the "success story" of the hard-top convertible which has been ascribed 
to the fact that a sedan is supposed to represent a wife— safe, sedate; while a 
convertible represents a mistress— dangerous, exciting; hence, the hard-top con- 
vertible should be doubly popular because it appeals to a man through two 
motivations. But Bauer points out that this just wouldn't be good psychology; 
to appeal to these conflicting motives could disturb a man, not lead him to 
buy. 17 In other words, the hard-top convertible became a success for more 
utilitarian reasons, such as being a very practical combination of attractive 
style and safety. 

In such a case, you certainly do not aim at the emotional motive. It is in 
your favor, so why tamper with it? Why run the danger of making the pros- 
pect aware of his illogicality (or wickedness, if in fact a socially unapproved 
sexual drive is operative here), stirring up a conflict in his mind, and actually 
causing resistance? You just give him a strong rational reason to "latch onto" 
—service, trade-in, or sunshine and health— and free his already existing emo- 
tional urge to lead him to it. 

C. If, however, there is a strong negative reason present, the problem be- 
comes more subtle. Go back a few years in imagination to the days when the 
automatic washing machine was first introduced; and take the case of a 
woman who wants to buy one for the simple personal reason that it will 
make life easier for her, but for whom this is not a self-approved reason. 
Moreover, such motivation is sharply in conflict with her unconscious picture 
of herself as a martyr to drudgery. She hesitates to ask for the product— that 
is, to ask her husband for it and/or even to ask herself for it; hence resistance. 

There are two possible courses of action here: (1) You can present a 
rational reason— such as "washes clothes cleaner" or "less harmful to fabrics" 
—which will liberate her desire for a joyful Monday, and off she will go (with 
her husband) to buy it. This is just what happened, and it happened quickly. 



17 Ramond A. Bauer, op. cit., p. 108. 



EXPANDING DEMAND 247 

(2) You also can try to change the non-self-approved to self-approved— in 
other words, to make it rational. This is a longer, slower process, yet by now 
hundreds of advertisers promoting hundreds of appliances have actually made 
labor saving for the housewife quite respectable, that is, in line with group and 
social norms. 

D. The difficulty is compounded when there is no strong motivation 
already existent and working for you. Negative reasons become correspondingly 
more significant. If the resistance is on the rational side, it can usually be met 
rationally. Most rational objections, if anticipated and met head-on, can be 
turned to advantage by a good salesman operating on the rational level (who 
gains the added emotional plausibility of presenting both sides of the question, 
as noted earlier). 

This is particularly true in business selling. In the case of a storekeeper 
hesitating to buy because he has a heavy inventory, the strategy is to show him 
that the turnover rate of the new product will decrease his investment per 
dollar of sales. Or, in the case of industrial equipment purchases, the quality 
of service features that add to price can almost always be translated into long- 
run dollar savings. 

E. The toughest situation of all is where the cause of resistance is emo- 
tional. Usually, it will be expressed rationally, and no amount of meeting the 
prospect on this ground will change the picture. Even if you win the argument, 
he simply will shift from one meaningless position to another. 

The surest course is, again, the long-run one of changing the irrationality 
of the resistance to rationality— making it socially acceptable, and thus self- 
approved. Take this situation: people may hesitate to make more long-distance 
telephone calls because of a feeling of uneasiness engendered by association 
with sickness, death, emergency, delayed arrival, and other unpleasant news 
conveyed by such calls— although they usually explain their reluctance on the 
ground of expensiveness. This may be one of those situations where emotional 
resistance is unconscious just because it is irrational and hence at odds with a 
person's picture of himself. 

But the telephone companies, by their advertising over a period of time, 
have begun to make the telephoning of friends and members of the family 
appear an acceptable thing to do— just witness the ease with which today's 
young people pick up the phone and place a call (more often than not, col- 
lect). For the new generation it is rational; it fits the norms of their group. 

Perhaps this telephone advertising would have been still more effective if 
all mention of cost had been omitted. Even a claim that "it costs less than 
you think" calls attention to the fact that there is such a convenient rational 
explanation for having an emotional feeling against long-distance calls. If 
so, this is a situation where only a long-range rational approach will fit into a 
good strategy, and an immediate rational appeal may actually reduce effective- 
ness, whereas an immediate but different emotional appeal would be appro- 



248 EXPANDING DEMAND 

priate— for example, the warmness and intimacy of people talking together— 
so as to supplant the negative emotional feeling and give the longer-run 
rationalizing process a chance to go to work. 

In all these cases, the moral is obvious. Whether or not the selling strategy- 
incorporates an emotional appeal, it is necessary to know when emotional 
motivations are at work for you or against you, in order to select the proper 
approach. In other words, here is one more demonstration that motivation 
research does provide a valuable service when used for diagnostic purposes. 
This is not meant to deprecate the role of emotions in people's attitudes, but 
simply to confirm the fact that when it comes to the strategy of bringing about 
change, even in emotional attitudes, the rational approach or goal is indispensable. 
The rational approach helps on the selling end as well as on the buying end 
of the transaction. Think of a salesman. (Advertisers could sharpen their own 
efforts sometimes if they, too, assumed the salesmen's role.) The rational approach 
is more simple for a salesman to understand, more natural for him to plan, more 
direct for him to follow. Further, by concentrating on one goal for the prospect 
and moving toward it, purposely, he is likely to end up with an attractive presenta- 
tion. Since he is thinking in terms of a need or want, rather than a product feature 
or advantage, he is necessarily concentrating on the prospect as a human being 
with thoughts and feelings, and will therefore instinctively tend to act so as to 
satisfy the prospect's motivations. 

The fact that the goal is a rational need or want is also helpful in that, since 
it is acceptable, it gives the salesman the self-respect and confidence he must have 
if he is to keep on selling. Since it makes sense, he does not become embarrassed 
or confused. Not being a trained psychologist, he can be very wrong if he tries to 
figure out all the nuances of emotional motivations; and, even if by chance he is 
right, he will act self-consciously, and spoil the sincerity that is so essential 
for good low-pressure selling. 

But there is an even more fundamental reason for depending on rational 
appeals if one is aiming for a plural market— as most sellers are, whether it be a 
specialized group or the mass market. Everybody's self-images or bundles of motives 
are a little different, and it is necessary to choose an appeal that will best cover 
the particular market being sought— that will have great positive strength for most 
of the individuals, moderate to little weight for the balance (it is too much to ex- 
pect 100 per cent aim), and little or no deterring effect on any. For these purposes, 
a rational motivation usually serves best. Rational reasons are necessarily shared 
by a large part of the market, since they reflect the norms and customs of society. 
By the same token, they also are unlikely to restrict the market by any deterrent of 
embarrassment or confusion. 

Open Selling 

Actually, there is much to be said for simply stepping up the degree of 
selling effort in many situations— for selecting the marketing mix element that 



EXPANDING DEMAND 249 

is appropriate, and backing it with more money, more effort, and more enthusiasm. 
(There seems to be a mistaken notion that low-pressure selling means that the 
salesman mustn't even "let on" that he thinks his product is better, let alone 
wonderful; this is almost as serious as the notion that low-pressure means no- 
pressure — which is ridiculous, since then it stops being selling at all.) Indeed, 
selling usually gains from not being disguised. 

People are hard to fool— increasingly so. But there is not just the danger 
that sly selling will be detected and boomerang— thus ruining that particular at- 
tempt and impairing all other, more honest attemps. Rather, the trouble is that 
such selling may be inefficient because it does not make use of the positive 
effect of openly helping people to buy. 

Certainly, any apprehension on the part of buyers that someone is trying 
to sell them something without letting them know it, and without giving them 
a fair chance to make their own buying decisions, will create almost insurmount- 
able resistance. By the same token, selling that does not try to hide itself or 
pretend to be something different is reassuring— so long, of course, as it does not 
go over to the opposite extreme of pushing people^ to buy through blatant high- 
pressure techniques, which, again, will make them feel they do not have a chance 
to make their own decision on the rational basis that is so necessary to them. 
There is as much difference between high-pressure selling and good low-pressure 
selling— open, purposeful, low-pressure selling— as there is between such open 
selling and hidden selling. 

People expect that a seller who has something to sell will want to sell it; 
that is rooted in our culture. Further, they respect sellers who have enthusiasm 
for their products; that, too, is natural. Indeed it is cause for alarm, and thus for 
resistance, if a seller apparently is not convinced he has a good product. And since 
the essence of the low-pressure technique is to present the product as the solution 
to some problem or need of the buyer, the more purposeful the approach, the 
more the buyer feels he counts. 

It is criminal waste of the buyer's own self-interest not to use it to lead 
him to the product or the brand. Actually, he is likely to be happier if he is so 
led. The more he is made to want it (short of misrepresentation and disillusion 
after the sale), the happier he will be when he buys. Again and again, a hopeless 
seller has lost sales while a hopeful buyer waited — and, in fact, wanted— to be 
persuaded to buy. For example: 

A middle-aged couple came into an appliance store and asked about 
a color TV set. In general they showed, by the models they looked at and 
the remarks they made, that, without realizing it, they had already made up 
their minds to spend a large amount of money. But every time they raised 
a question about price, the salesman quickly exhibited a cheaper model, not 
realizing that what they were asking for was not a better bargain but some 
assurance that there was a rational reason for them to pay the amount in 
question. 



250 



EXPANDING DEMAND 



At least three times they were on the verge of saying, "We'll take it." 
At least three times the salesman lost them by not trying to sell them. They 
left the store without buying. As they went out, they said— and by this time 
perhaps thought— that they could not afford the money for one of the small 
sets, yet it was obvious that they were completely unhappy because they had 
not been sold the big set they really wanted. 18 

This idea also applies to advertising. Dull advertising can actually cause ill- 
will and evoke criticism, just because it doesn't build enough want to make the 
promise of satisfaction interesting and worthwhile. The marketer needs to be 
exciting; there is a line between being dramatic and being misleading, but the 
marketer must not be afraid of getting close to this line at the same time that 
he respects the importance of not going over it. 

At the other extreme, off-beat copy themes which deprecate the product have 
proved singularly ineffective. Sometimes they achieve some temporary success 
just because they attract attention by their oddity, their self-ridicule, but usually 
people resent the lack of dignity; sellers who take justifiable pride in their 
products could not possibly talk or think like that. (This does not refer to 
cartoons that laugh with the product, rather than at it, or to other unusual modes 
of advertising used primarily for getting attention or for change of pace; after 
all, one basic function of advertising is to keep the name of the product before 
the public) . 

If anything, the need for open selling is even greater in advertising, which 
has to depend on fleeting impressions and the competition of a multitude of sights 
and sounds. People have developed a defense mechanism against even noticing 
ads which do not bear on some problem or want they already have in some 
degree. Today, a picture of a refrigerator and a headline of a new feature will 
attract the attention of those who are refrigerator-minded, while the John-loves- 
Mary theme, where the refrigerator only comes in subsequently, is literally 
passed by. 19 It makes sense that people will see or hear the message that speaks 
to them in the terms of their own specific everyday interests— and one of the 
most important of these is the buying of specific products or services. So why not 
at least try to make a definite selling impact? 

The utmost in waste of good dollars and glossy paper would seem to be the 
nonselling that is characteristic of much advertising of industrial products or 
services. It may be true that here the burden is on the salesman in the field; but 
even in the secondary role of seeking inquiries or paving the way for the salesman, 
the advertising message will come through stronger if it has at least enough "sell" 
to be pointed specifically and purposefully to the prospective buyer's problems. 



18 Reported in "Northeastern Distributors, Inc., Recording — Part II," a case prepared and 
copyrighted, 1954, by Harbridge House, Inc., Cambridge, Massachusetts. 

19 See Richard D. Crisp, 'Thinking Ahead: Advertising Research," Harvard Business Re- 
view, March-April 1953, p. 30. 






EXPANDING DEMAND 251 

Why waste the opportunity to do some selling, when that must have more mean- 
ing to the buyer than any dull generalization or "institutional" portrait of the 
company? 

This means the opposite of simply augmenting effort by increasing expendi- 
tures and repeating the same message. There is increasing evidence that repetition, 
if carried too far, has diminishing utility, and may even end up with a negative 
effect through boredom or annoyance. 20 But even if this were not the case, it 
would be foolish to let slip the opportunity to enrich the basic appeal by ex- 
pressing it in several forms, to vary the message enough to appeal especially to 
different segments of the market, and still get the advantage of frequency, con- 
tinuity, and cumulation (the slogan or logotype can be repeated, of course, within 
the changing context of the ad or selling presentation). Indeed, what is being 
advocated here is consideration of the potential in more selling just so a better 
job can be done in the light of the particular product-market situation. 

In sum, most people have healthy minds; they like to buy. There is no need 
to dig into their subconscious to free them from blocks and tensions before 
they buy. (In fact, such an effort may be dangerous— may stir up conflict without 
enough time or skill to remove it— something even psychiatrists worry about.) 
Emotional resistances, even if in the subconscious, can be handled by the healthy 
prospect himself— and will be, if he is provided with a sufficient rational motive 
as incentive, along with whatever emotional appeal is appropriate. 

But selling is not just satisfying present wants or playing up to old desires. 
Selling is the process of expanding demand. Not only is this what keeps our 
economy dynamic, but, to the extent that it gets consumers to consider the 
pursuit of noneconomic values as approved and rational, it is also building people's 
potential for a higher, less materialistic way of life. 

The danger of consumer motivation research is that business may lean back 
and leave everything to "scientific" techniques which are not meant to do the 
whole job, and forget to go out and sell. As far as getting buying action is con- 
cerned, the actual psychological subtleties may just be too tenuous for the hit- 
and-run of daily life. But we can use our understanding of how people buy to 
build a lot of little strategies for persuading people to want specific products and 
services, and one grand strategy of giving people continually bigger and better 
goals for themselves. 

In either case, there is a big, uncomplex psychology at work— the interaction 
of enthusiastic seller and eager buyer, out in the open where they belong. And 
whatever emotional forces there are will be released if they are in your favor, 
contained if they are against you. Such selling is not manipulating people behind 



20 Cf. Rosser Reeves, Reality in Advertising (New York, Alfred A. Knopf, Inc., 1961), p. 32, 
for the argument that repetition is desirable, and, for the other side of the question, Carl I. 
Hovland, Irving L. Janis, and Harold H. Kelly, Communication and Persuasion (New Haven, 
Yale University Press, 1953), p. 247. 



252 EXPANDING DEMAND 

their backs; it is giving them rational motives for doing what is in their own 
best interests as individuals and as society-at-large. 

Now let us go on to look at the various aspects of promotion from the 
management point of view. 

The Selling Job 

It should be clear, from the foregoing discussion, and also from some of the 
cases attached to this section, that effective selling requires a much higher 
degree of intelligence than is commonly supposed. Mental agility and sensitivity 
are particularly important in a low-pressure approach, where the salesman, often 
beginning with a question (like "Have you had any problems with truck break- 
downs on the road?"), must be able to follow the lead he gets in response and 
develop the conversation until his product or service (say, motor oil) comes into 
it naturally. 

Also— and this point often gets overlooked— a good low-pressure approach 
requires considerable planning. Remember, it is not just going in to see a prospect, 
saying things to please him, reciting a few facts, and then hopefully letting the 
interview develop as it will. It requires a great deal of preparation to be able to 
take advantage intelligently of whatever opening the prospect gives you, and this 
cannot be done unless the various alternatives of what he may say are anticipated 
and provided for. Also, analysis of the individual prospect, or class of prospects, 
is needed, followed by planning to fit the gambit to the particular situation so 
that the more favorable alternatives are more likely to be the ones that do de- 
velop. For example, in the case of anticipated resistance, one planned approach 
(rather than starting with a general, provocative question) is to deliberately raise 
the issues on which you expect arguments from the prospect before he brings 
them up, to meet them head on, and then to turn them to advantage. Thus, taking 
an anticipated price objection (in oversimplified form): 

"The line of insulcapacitators I represent is the highest priced in the 

industry 

"Yes, and they're high-priced on purpose, because we are only interested 
in selling to those who prize trouble-free service more than a few pennies' 

bargain 

"Well, it costs money to build into the product the kind of extra rugged- 
ness and dependability that ours has. Here, let me show you the specifica- 
tions. .. ." 

This is an example involving an industrial product, intended to illustrate the 
fact that, although selling would seem to have less scope in the industrial area, 
it actually has, if anything, more. There is no gainsaying that purchases of things 
like machine tools and raw materials, being made for business purposes, will 
largely be guided by consideration of impersonal factors like specifications. On 
this basis, competition would have to be primarily in terms of product develop- 



EXPANDING DEMAND 253 

ment and engineering know-how— in short, technical service. But it is the very 
fact that all companies are competing this way that leaves the salesman with no 
overt advantage for his product— which may help to explain why so many indus- 
tries, in the absence of effective selling, allow reciprocity to determine purchase de- 
cisions. And, for the same reason, the better seller of two companies, everything 
else being dully equal, is bound to get the business. 

In one respect, consumers are becoming more like industrial buyers. An 
increasing number of consumers put an obvious premium on price, judging 
by the way they patronize the discount houses to buy electrical appliances or 
shop around for trade-ins on cars. But if you accept this as inevitable— as a num- 
ber of manufacturers do, even though they pretend to fight it— then you have lost 
control over the movement of your product. Your advertising serves mostly to 
make your dealers prize their franchises, and the best you can do is to load them 
so heavily that, in self-defense, they have to try to move your product, which 
ends up by causing even more price cutting. 

At the same time, other manufacturers are using advertising to do practically 
the whole job of getting their products into the consumers' hands— cigarettes, 
soaps, cosmetics, cleansers, beverages, certain fast-moving food items, and so on. 
With media like TV exerting more powerful impact and with consumers guided 
more by their own impulses (self-service), it is increasingly possible to buy a 
monopoly position on the market— until somebody else comes along and outbids 
you for it with a larger advertising budget. Brand loyalty has been shown not to 
be the simple, steadfast thing it once was considered; it is real and influential, 
indeed more so than ever, but it is impermanent, shifting, always up for auction 
—and expensive. 21 

This is not to suggest that there is anything wrong with employing service, 
price, or advertising to move products. These, along with personal selling, are all 
elements of the marketing mix, and different strategies will put different weights 
on them— and succeed if there is a large enough percentage of the market that 
wants that particular combination. But it also is true that putting reliance on one 
element does not require ignoring the other elements— particularly selling. 

There is much to be said for selling being more important in inverse propor- 
tion to its weight in the marketing mix. To the extent that the movement of 
your goods depends on forces not directly within your control— the industrial en- 
gineer's specifications, the retailers' pricing, the advertising agency's skill— the 
area where you can take energizing action becomes smaller, but what you do 
there also has increasing leverage. Your competitors are subject to the same 
conditions; and the better selling you do, the more you will pull ahead of them. 

Furthermore, even where personal selling does not have the prime or ultimate 
role in moving goods, it does come in— or should come in— at crucial points. The 
manufacturer of cereals still needs to persuade the supermarket to give his product 



21 See the appendix to Case 5-2. 



IP 



- 5 H 
* v § 






to ^ 



00 



CD CO 
»-H rH 



■^ id 

id ■<*< 



s 



IP » S 



c^ 



rH CO 
t-' CO 



w w 



§ 


o 


© 


o 


© 


o 


© 


© 


O 


o 


o 


o 


o 


5 s 


o 


o 


o 


■<* 


CO 


o 


o 


o 


o 


o 


CM 


o 


CD 


N 


"tf 


CD 


oo 


©^ 


CM 


-* 


o 


CO 


CO 


CM 


^s -M 




























C-" 


c-~ 


"* 


CO 


Oi 


a> 


CD 


in 


ID 


V 


"«* 


"* 


rH 


1-* 


T— 1 


i-H 


















s 


V* 

























o 



o o id o o 

O O CO C- o 
O ID CM rH t- 



o 


© 


o 


o 


ID 


O 


ID 


o 


o 


ID 


o 


CO 


o> 


"^ 


"tf 


Oi 


CO 


■"*■ 


CM 


H 





CM i-l rH 



s 8 

■g IP 



o o o o o o 

o o o o o o 

O O^ O CO CO o 

^ CO CM rH rH r-T 



o 


© 


o 


O 


o 


o 


o 


o 


o 


o 


o 


CM 


CO 


o 


oo 


co 


ID 


">* 


CO 


CM 


CM 



<*) n 










^3 <*> 2 

.2 ^"-i O 


O 


o 


o 


ID 


-2 c^o cj 


ID 


o 


o 


CO 


*» ^ **» »> 


CO 


o 


ID 


00 


ID 


ID 


•>* 


CO 




#y 








■ft. <o 











o 


o 


o 


© 


o 


ID 


O 


ID 


o 


o 


o 


ID 


o 


ID 


ID 


"^ 


c- 


CM 


ID 


rH 


co 


ID 


"^ 


CO 



CM rH rH 



« 






* 




<fc 


"•o 




S 




5 


a 


jtf 


» 


g 


o 


CO 


co 


*o 


co 




fc 


5? 




r. 


cu 

s 




•o 


co 




s 


CD 




,3 


a 




* 


CO 



o o o o o 
o o o o o 

O O rH CM O 



o o o 
o o o 

CO o c- 



o 


o 


O 


© 


ID 


o 


© 


o 


ID 


O 


o 


ID 


o 


00 


CO 


ID 


"^ 


CO 


CM 


CM 


CM 


^-* 


y-* 

















254 



EXPANDING DEMAND 255 

Notes to Table 4-B 

a 35% X sales, less fixed cost of $600,000 up to $8,000,000 sales; $800,000 up to 
$12,000,000 sales; $1,100,000 up to $16,000,000 sales; $1,300,000 up to $19,000,000 sales; 
$1,500,000 up to $21,000,000 sales; $1,900,000 up to $23,000,000 sales; and $2,400,000 
up to $25,000,000 sales. 

b Estimated at $20,000 per man. 

c Column III minus Column IV. 

d Working capital + present plant investment + incremental investment, where working 
capital = 40% X III; present plant investment = $3,000,000; and incremental invest- 
ment of $2,000,000 is needed above $8,000,000 sales, an additional $3,000,000 above 
$12,000,000 sales, and an additional $2,000,000 above $16,000,000 sales. 

e V-MI. 

'V-*-VI. 

SOURCE: Adapted from Walter Semlow, "How Many Salesmen Do You Need?" 
Harvard Business Review, May-June 1959, p. 130. 

advantageous display. The manufacturer of automobiles still needs to convince 
his dealers to follow up sales leads. The manufacturer of industrial equipment 
still needs his engineer to present specifications in a form that will answer the 
buyer's unasked questions before they turn into objections. 

SALES MANAGEMENT 

Personal selling still accounts for the largest part (about 80 per cent) of 
marketing costs (in turn, averaging 50 per cent of the total dollars paid by con- 
sumers and users of goods and services); and presents management with a series 
of constant, demanding problems that have to be solved, not only for their own 
sake, but also because of their effect on the rest of the organization. Let us 
look at some of these sales management problems. Though they are often 
assumed to involve personnel matters almost exclusively, there are a number 
of points at which they tie in with strategic marketing considerations, and it is 
wise to examine them from the same analytical viewpoint. 

Number of Salesmen 

Take first the problem of how many salesmen to have. There are a number 
of pieces of information that are needed: 

(1) What is the potential sales volume, in total and by territories? The 
total figure is indicated by published figures, or is estimated by the com- 
pany. The territorial figures are derived by multiplying the national figure 
by the proportional index figure for each territory. The index may come from 
a recognized source like Sales Management magazine's annual "Survey of 
Buying Power," or, if the company's product involves some special factor or 
factors more appropriate than general purchasing power, from the company's 
own analysis of what each territory ought to produce as a proportion of the 
total. 

(2) What per cent of potential in a given territory can be achieved 
by how many salesmen? That is, one man might be able to get 40 per cent, 



256 EXPANDING DEMAND 

two men 70 per cent, three men 80 per cent, and so on. This is the usual 
kind of pattern, since one salesman can skim the cream (say, the best of 
the 20 per cent or 30 per cent of the customers who account for 70 per cent 
or 80 per cent of the total business), and each succeeding salesman brings 
in a smaller increment of sales. The data here can come from an analysis 
of the percentage of potential currently being secured by different numbers 
of salesmen in territories of different potentials (omitting territories where 
there are unusual situations and smoothing out the irregularities by statistical 
or judgmental methods). However, the data can be refined considerably by 
experimentally increasing or decreasing the number of salesmen in a territory 
for just this purpose, and observing the results; indeed, this is highly de- 
sirable in view of the danger that existing arrangements may not cover the 
spectrum of possibilities and thus may tend to perpetuate existing inadequa- 
cies or excesses. 

(3) What does it cost to maintain a salesman in the field, including 
the amortization of the company's investment in his training, and an alloca- 
tion for the cost of turnover of salesmen generally? 

(4) What is the company's operating profit before selling costs for 
different volumes of sales? Ordinarily, the profit figure will rise faster than 
sales volume because of the large element of fixed manufacturing costs. 

With this information, it then becomes possible to make an analysis such 
as that shown in Table 4-B (it might be even more helpful to plot the figures 
and draw curves graphically, too). Somewhere around 100 men is indicated as 
the optimum number of salesmen in terms of maximum dollar profit, or 65 men 
in terms of maximum return on investment. (The latter criterion, as noted earlier, 
is coming into increasing use as a benchmark for decision making and appraisal 
of performance.) Such calculations are only as good as the figures fed into them, 
and must be tempered by judgment and used with caution. Often, management 
will want to take special steps, such as putting fewer men in a territory where 
the company is at a competitive disadvantage because of special circumstances. 
Analytical exercises like this may be useful sometimes just by altering managers 
to the fact that in making such moves they are departing from the course that 
general conditions would call for. Even when the results of the analysis are ex- 
tremely rough (so long as the roughness is recognized), the discipline of trying 
to think through the problem in quantitative terms may turn out to be extremely 
helpful. 



Selection of Salesmen 

Usually more important than the number of salesmen is the problem of 
having the right kind of salesmen. The top man on a sales force often brings in 
three or four times as much business as the poorest performer. 



EXPANDING DEMAND 257 

In many sales forces the difference between the top producers and the "also 
rans" is not in drive, social ease, empathy, technical knowledge, facility of ex- 
pression, and so on (no matter how helpful or indispensable any or all of these 
may be), but in the degree to which the better men (a) plan their time, their 
analyses of prospects, and their presentations, in advance, and (b) make self- 
analyses of their own action and their prospects' reactions, after the event. 

Certainly the trend is in the direction of more planning, more individualiz- 
ing to the situation. Eugene B. Mapel, Vice President and Director of Marketing 
Services, Chase Manhattan Bank, has pointed out: 

As the complexity of markets and products increases, the effect of each 
marketing decision and each sale becomes more intimately interrelated with 
all others. Managements try to establish policy at headquarters that will 
guide salesmen, but it's getting increasingly difficult to foresee all the ramifica- 
tions of each decision. More and more frequently the individual salesman 
must know when not to sell. He must think through such factors as the 
prestige impact of one sale on other customers and prospects. He must 
ask himself, "If I sell 'A,' what effect will this have on the firm's prospects 
of selling 'B,' who could handle a greater volume, or *C,' who would enhance 
our product's prestige?". . . More and more salesmen are being called upon 
to make just such decisions. 22 

Some selling doesn't support the luxury of carefully planned, individualized 
approaches. To the extent, and only to the extent, that the salesman can be 
given a spelled-out guide to what he should say or do, or even a "canned" 
presentation to be memorized and repeated in all situations, is it possible to settle 
for more persistence and less intelligence. 

Psychological aptitude tests have been found very useful in screening sales- 
men; they serve better to screen out men who are going to be duds than of funnel- 
ing in those who will be pre-eminent successes. But they have to be suited to the 
particular selling task. Further, one of the main reasons the tests are useful is that 
they cause all concerned to be more impersonal, hence more objective, about 
who is going to be hired as salesmen. 

The need for objectivity is important. To illustrate: 

In one large life insurance company, the same field office supervisors 
who had exercised poor judgment in recommending the hiring of candidates 
that they had interviewed for sales positions made an almost perfect record 
at a training session, in correctly judging the future performance of can- 
didates whose backgrounds and qualifications they only saw in written form. 
They had no more actual information, but they avoided being swayed by 
personal impressions. 



22 Proceedings of the Boston Conference on Distribution (Boston, International Marketing 
Institute, 1961), p. 37. 



258 EXPANDING DEMAND 

What happened was that the field office supervisors were encouraged, in 
their training sessions, to analyze what characteristics the job required and then 
to watch for the items on the candidate's application paper (including a credit 
survey and appraisal by former employers). In this particular instance, there 
were several items that were of special significance: 

The candidate's statement of what he expected from the job. How 
realistic was he, in the light of the day-in-and-day-out need for persistence, 
capacity to take rebuffs, and so on; and how ambitious was he for financial 
success in the light of the fact that insurance selling requires a good deal 
of drive to keep a man at it? 

The candidate's indications of intelligence, in the light of the need 
to be able to understand and explain fairly intricate contracts, but not to be 
so intellectual as to look down on the job. 

The candidate's previous employment record. How long did he stay with 
each job, and did his previous work involve relations with people (not 
necessarily selling) in the light of the combined need for stability and for 
ability to make contacts with people easily? 

The candidate's memberships in social clubs or organizations, again in 
the light of the importance of personal relations. This does not mean that he 
must be an all-out extrovert— the jovial, back-slapping type that is often 
pictured in popular literature. In fact, that type of person may turn out to be 
unsuccessful for the very reason that he is not sufficiently sensitive to other 
people's feelings. 

In general, the characteristics of the job that are likely to be important, 
either in requiring or dispensing with particular traits like the above on the part 
of the salesmen, are: (a) how directly the man is supervised, (b) what kind of 
knowledge the product involves, (c) with how much flexibility and /or tact the 
typical customer must be approached and handled, and (d) whether the major 
content of the job is new calls or repeat calls. When in doubt, it is probably 
better to err on the side of intelligence, since an intelligent man can learn, can 
acquire most of the other skills, and, if not tactful, can discipline himself to be 
so; but a man who is tactful because he is naturally empathic cannot train himself 
to be intelligent if he not already blessed with the requisite amount of brain 
power. 

The training program, the compensation plan, and the organization for field 
supervision all must reflect the same requirements. Thus, in training, to the 
extent that flexibility is required in fitting the sales presentation to the particular 
situation, there is both need and opportunity for teaching the salesmen how to 
analyze a prospect and compose an approach that will create the desired effect; 
while, in opposite situations, he can be drilled in routine openings and stock 
responses. 



EXPANDING DEMAND 259 



Sales Compensation 

In compensation, there are two questions: how much to pay, and in what 
form to pay it. 

The question of amount depends only in part on the requirements of the 
job itself. It also depends on what the company can afford to pay in terms of its 
margin on each unit of sales (that is, the difference between the selling price 
and the manufacturing or wholesale cost), though here, just as in advertising, 
spending more may actually reduce selling costs per unit or make a larger 
margin possible. That is, the higher-paid salesman may be enough better so that 
he sells more per dollar of compensation, or is able to sell the product success- 
fully at a higher price. 

Note, however, that what the company can afford to pay does not necessarily 
mean that it should pay that much if the job does not require it. The fact may 
be that the company can evolve a more efficient marketing mix by putting the 
additional money into advertising, so that the product doesn't actually need as 
much good personal selling. 

Another factor that has an influence on how much to pay is competition— 
that is, what other companies are paying salesmen of similar caliber. But, once 
again, everything starts with the requirements of the job itself: What kind of man 
do we need to handle this selling job? 

As for the form of compensation, there are two main varieties: (1) fixed 
(straight salary) and (2) incentive (percentage commission or any other form 
of paying in proportion to the sales that a man, or group of men, makes). Most 
plans involve some combination of the two. Thus, the sales force on straight 
salary may receive annual bonuses based on their sales performance, or a per- 
centage for all sales beyond a certain quota; while in the sales force on commis- 
sion each individual may be guaranteed a certain minimum, or may have a 
"drawing account" from which he receives money in advance of sales, to be 
covered by his subsequent commissions, thus regularizing his pay and making it 
more like a salary. 

Both elements have some advantage. The more commission-like the plan is, 
the greater is the incentive to work hard and push for immediate sales; but the 
more salary-like it is, the better the salesmen will follow directions and engage 
in activities (such as working on new accounts or helping his customers with 
their problems) which are likely to build sales in the future. 

The fact that these two advantages are both desirable and both in conflict 
with each other means that there can be no such thing as a perfect compensation 
plan. (This helps to explain a phenomenon that takes place in many companies: 
periodic shifting of the salesmen's compensation plan, the first time toward 
salary, the next time toward commission, then back toward salary, and so on.) 



260 EXPANDING DEMAND 

Here, too, the nature of the job and the kind of man required are crucial 
in coming to a decision as to just where along the range from salary to com- 
mission the plan should be. While the various factors that push a plan in the 
one direction or the other are often overlapping and even opposed, the marketing 
manager who is analyzing the situation in his company will find it useful to think 
of the different elements that he must weigh and combine, or put together 
in the best possible compromise. Thus: 

Factors Pushing Toward Salary Side 

Low level of compensation, hence need of assurance of regular pay. 

High content of nonselling activities, that is, special assignments or 
duties that do not result in immediate sales. 

Low caliber of salesmen and/or absence of close supervision, so salesmen 
must have prescribed routines (which it is difficult to get them to follow, 
if they have their eyes on commissions). 

Fluctuating and erratic sales, hence need for stabilizing the salesman's 
earnings (an extreme example would be selling aircraft to the government or 
to airlines). 

Products that must be sold slowly and without push, or relations that 
must be maintained in a nonaggressive manner. 

Situations where a number of salesmen and other company people must 
cooperate to make the sale, or where, for any reason, it is difficult to allocate 
the credit for making the sale. 

Factors Pushing Toward Commission Side 

High level of compensation, so salesman has some leeway in his family 
budgeting, some credit to fall back on in anticipation of increased future 
earnings. 

Situations where the salesman needs great incentive effect because drive 
and initiative are essential to success. 

Intelligent and technologically knowledgeable salesmen, who will want 
to gain from their own self-starting and planning, and who are of high 
enough caliber so the company can afford to give them this privilege. 

Situations where the compensation plan is to be used as a mechanistic 
substitute for supervision; that is, with different quotas and different per- 
centage commission for different products, different customers, or different 
kinds of sales (for example, a higher commission on initial sales to new 
accounts). 

As suggested by that last point, there are also a number of forms in which 
commissions can be paid. One of the prime questions is whether the percentage 
should be constant, no matter what the sales volume, or should decline with 



EXPANDING DEMAND 261 

increasing volume (or on sales volume beyond a certain amount), or should be 
higher with increased sales volume. Again, management must weigh the pros 
and cons of the different possibilities. The constant rate is clear and simple. 
The declining rate, as long as it is graduated in such a way that the salesman still 
receives more money for more sales at any point, does give the company the 
advantage of sharing in the salesman's productivity. The rising rate, for com- 
panies whose profits go up rapidly with increased volume (for example, where 
capital costs are high), has the merit of keeping the salesman always hustling for 
more. 

Note, again, the impossibility of devising a perfect salesman's compensation 
plan because the objectives of various alternatives are all partly good, partly bad. 
Thus, the more simple the plan, the less it will direct and lead in the areas of 
activity designed for long-run effect or indeed for anything except immediate 
sales; conversely, the more elaborate and tailored to various specific activities, the 
more blunted is the plan's sheer incentive effect. 

Take the matter of quotas for salesmen— the amount in dollars or units (or, 
even in some cases, margin or profit) which they are expected to sell in a given 
period of time. Whether tied in with the sales compensation plan or simply 
used as a basis for judging performance and annual salary review, quotas can be 
helpful. But they involve a number of messy imperfections, represented by such 
questions as: "Do we set quotas that can easily be made, or made through 
moderate effort, or that are so tough that only all-out effort by the best men can 
accomplish them?" The obvious difficulty is to call forth extra effort without 
going so far as to have an offsetting discouragement factor. 

Another difficulty is to arrive at quotas in such a way that they are under- 
standable by the salesman, and the salesmen feel they are just and fair in line 
with the situations in their individual territories, the condition of the economy 
as a whole, and so on. Finally, there is the difficulty of setting quotas, and tying 
them into a compensation plan, when the selling activity is a group effort; one 
solution here is a group quota, but this, in turn, does not have as strong an 
incentive effect on the individuals in the group, and there is likely to be per- 
sonnel trouble if some are working harder than others. 

But, above and beyond all such difficulties, is the fact that no plan is 
perfect in the sense of being a substitute for intelligent, understanding sales 
management, particularly in the field. 



Field Supervision 

The role of field sales management— branch manager, district manager, or 
whatever title it may carry— is crucial. 23 A number of studies have demonstrated 



23 See "Field Sales Management — Weak Link or Strong Link in the Selling Chain," Acme 
Reporter, 1960, No. 1. 



262 EXPANDING DEMAND 

that the largest determinant between success and failure for new salesmen is the 
quality of supervision they receive. A weak salesman with a strong manager may 
actually have as much chance of success as a strong salesman with a weak man- 
ager. And salesmen of equal ability will have from two to five times as much 
chance of success under a good manager. 

This is not to say that the quality of field supervision is equally important 
in all companies; but it never is unimportant. Much depends, once again, on the 
caliber of salesman, the distance from the home office, the kind of products sold 
(and hence the kind of customers), the nature of the company's compensation 
plan, and so on. Similarly, the responsibilities of the field sales manager will tend 
to vary with the situation. 

Let us look at some of the differentiating factors, and then at some of the 
general principles that can be applied, in varying degrees, to all situations. 

A salesman works pretty much on his own. At best, he lacks the control 
and guidance that comes from direct contact with headquarters. He is separated, 
for periods of time, from the steadying influence of his home. He must con- 
tinually energize himself to start early, to tackle tough and perhaps hostile pros- 
pects, to cut wasted time, and, in general, to keep exerting effort when there is 
no immediate spur to do so. Also, very often, the nature of his work is intangible 
and indirect— building good will or laying the foundation for sales in the future— 
with no one to smile readily in approval of good work done (sales compensation 
cannot ever substitute for human warmth). 

Salesmen may act as if they want to be independent, but the fact is that 
they are more dependent on local managers, in actuality, than either they or 
their companies realize. Just because the life is essentially a lonely one, the 
encouragement and association with the firm's officialdom is important. They need 
to be sure they are being evaluated by someone who knows them; they need a 
person to give them advice or information in an unusual situation; they have 
to have a reliable source of facts on company policy. Like everyone else, they 
have their emotional ups and downs; personal problems and financial worries 
upset them. In most selling, the impression made by the man himself is a 
significant plus or minus in the closing of the deal; if the salesman is "off his 
oats," valuable orders may go by the board. In these situations, a man will be 
bolstered and guided by a sympathetic boss to whom he can turn. 

But there are differences of degree. The higher the caliber of the salesman, the 
less he needs someone else to do this for him; conversely, the lower his caliber, 
the more he needs it. Hence, in industries where the logics of the marketing op- 
eration call for men of less all-around maturity, there need to be more super- 
visors per salesman so they can be closer and more helpful. Or, to put it another 
way, in such situations the optimum mix is less money spent for salesmen, more 
money spent for field sales management. 

The opposite situation prevails when the nature of the product or the 



EXPANDING DEMAND 263 

customers is such that the salesman must be highly intelligent and responsible- 
situations where the content of the sales presentation is technical or engineering 
in nature, or where the sales unit is large in dollars and requires considerable 
preparation and negotiation. Then, whatever the kind of salesman required, the 
farther he is from headquarters (whether in sheer miles or in frequency of con- 
tact), the more need there is for someone to bridge the gap. 

As we have seen, there is often a tendency for a company to try to control 
the salesman through the sales compensation plan— sometimes in the form 
of direct incentive for more effort, sometimes through specific rewards for par- 
ticular kinds of activities. Any company, trying to decide what compensation plan 
will best motivate its salesmen, must also take into account what degree of field 
management it already has or is prepared to institute, or vice versa. The greater 
the degree of purposeful supervision, the less need there is to try to accomplish 
the same things through the more mechanistic means of a sales compensation 
plan. On the other hand, if the salesmen are such that they are good planners of 
their own time and effort, with due regard for the long run or whatever is in the 
company's over-all best interest, then a strong incentive plan may be more eco- 
nomical (that is, more effective for the same amount of expenditure) than a more 
intensive supervisory setup. 

In any event, the two means of motivating and controlling salesmen need 
to be considered together— compensation supplementing supervision, and vice 
versa. Failure to make the combination and to adjust the balance on this side or 
the other can be costly; and, from the point of view of profits, it is just as bad 
to have needless duplication (strong incentive pay when supervision is close, or 
strong supervision when compensation will secure the desired result) as it is 
not to have enough of either. 

SELECTION OF FIELD MANAGERS. In view of all this, marketing man- 
agers can well pay attention to the question of selecting the right kind of men for 
their companies. Once more, there is need of analyzing what the field manager's 
role should be in the particular company. 

What should the field manager's responsibilities be? Clearly, he has more 
to do when close supervision is called for— or, rather, he has more to do of a hands- 
holding or coaching nature; maybe training and development is a more com- 
prehensive way to describe it. Thus, the factors we have just discussed also affect 
the proportions of the elements that go into the mix of the field sales manager's 
job. But they do not change the one general fact that, in whatever proportions, 
that job is made up of two principal elements: (1) human relations activities, and 
(2) planning and decision-making activities. It is that second, more analytical 
element, which makes the manager's job so different from the salesman's. 

The salesman who has learned how to get along with people— to persuade 
them or to lead them to his way of thinking— already has some start on the 
human relations side of management. Yet even here there is a difference. A 



264 EXPANDING DEMAND 

manager must be more impartial and impersonal, since he deals with a number of 
people simultaneously rather than one at a time, and since he is concerned with 
helping his men to develop and grow, rather than with getting signatures on 
dotted lines. When, in addition, he must make decisions for the good of the 
company, instead of fighting for the customer's needs or desires, and plan for 
long-range goals instead of driving for this week's or this month's or even this 
year's sales, he indeed has to shift his mental gears. 

Hence, many companies have found— often to their sorrow— that to be a 
good salesman is not a guarantee of being a good manager. But the converse 
is not true; indeed, it is highly desirable that an effective field sales manager 
should also be a good salesman. There are two reasons for this: (1) the nature 
of the selling job is such that it is difficult to understand its hardships and its 
joys without having had some actual experience in it, although most salesmen do 
exaggerate their uniqueness; and (2) just because salesmen do feel this way, they 
are less likely to respect and have confidence in a manager who has had no 
selling experience. The second reason is, if anything, the more important. Just 
the same, most companies have found that, if they have to choose between a 
poor manager with selling experience and a good manager without it, they are 
better off to take the nonexperienced man. It is easier to learn what selling is, 
particularly if one is already equipped with management skills, than it is to 
learn to be a manager when one does not have the basic ability. 

The usual route to sales management is through selling. To get good 
salesmen it is often necessary to offer advancement to the management level as 
a goal; and, as just suggested, the previous experience is a highly desirable, if not 
completely indispensable, qualification. The question is how to select the right 
man. 

The first requirement is to have enough potential managers in the sales 
force to draw from— and not too many, for the morale of a sales force can be 
ruined if too many men become frustrated through a lack of advancement. A 
good rule of thumb, some companies have found, is to try to have not less 
than two nor more than three men for each foreseeable opening. This means that, 
in selecting salesmen, care should be taken to evaluate management potential as 
well as present sales ability, and to select the proper mix on this basis. But much 
more is required than to ask each applicant whether he wants to be a manager; 
any good salesman will answer "yes" if he thinks that is what you are looking 
for— and will know how to back it up convincingly. 

If psychological tests are being used as a screening device for hiring salesmen, 
they can be employed not just to eliminate men whose chances of becoming 
good salesmen are not worth the risk of training and "fielding" them, but also 
(with a moderate amount of supplementary effort) to distinguish between those 
with and without management potential. Thus, the company gathers information 
at the time of original employment to be used later when the management open- 
ing develops. But by this time the company has additional insight. It can take 



EXPANDING DEMAND 265 

the two or three men already planned on, plus maybe another one or two who 
have developed unexpectedly well, and, for each one, look at such things as: 

Does he have a good (but not necessarily the best) sales record? 

Has he been helpful to other salesmen or to the group as a whole? 

Are his interests in the company clear and direct? 

How good has he been at planning his time and selecting his prospects? 

What kinds of unusual things has he done like taking care of a par- 
ticular trouble-spot? 

Does he make snap judgments or reasoned decisions? 

Is he firm in his opinions, but willing to change them when shown to 
the contrary? 

Is he honest and personally above reproach? 

Once the man is selected, he still needs further management support, for 
he may well experience initial difficulties with the men under him; some of them 
may resent his being chosen— especially those who do not at heart really want 
to be managers themselves. But the company knows it has the right man if he 
turns out to be a happy person. For an individual with a liking for people and 
a taste for responsibility, there is no pleasanter position in life. And the salesmen 
whom he supervises feel this and respect him. Many managers who move up 
from the field to higher positions in sales management or general management at 
the home office look back with nostalgia to that part of their life. 

But more is required than attention to the criteria by which these men are 
chosen. If executives expect top-notch local supervision of sales forces, they have 
to take some positive steps to secure it. In training programs, new managers must 
be made aware of the dimensions of their new responsibilities, and given a clear 
picture of what the company expects from them. 

Furthermore, the job has to be equipped with the necessary authority. One 
can raise real questions, for instance, about the fairly common practice of taking 
the hiring function out of the hands of the field manager who must work with 
the salesmen being employed. Similarly, the home office should be aware of the 
impact on the local operation of sudden transfers which pull out the top per- 
formers for jobs elsewhere. This latter practice became so common in one firm 
that local managers, up against tough recruiting problems, purposely hired second- 
rate people to insure that they would have, at least, a stable corps with which to 
work. 

SPECIFIC RESPONSIBILITIES. By implication, we have already discussed 
the general responsibilities of a field sales manager— bolstering, training, guiding 
his men; planning things like the allocation of territory and distribution of special 
effort; making decisions in borderline and emergency cases or, depending on com- 
pany policy, in matters like special price quotations or delivery arrangements. 

Sometimes, also, he does sell too— handling particularly important or difficult 
accounts by himself or teaming up with one of his salesmen for the occasion. In 



266 EXPANDING DEMAND 

general, it can be said that there are two possible disadvantages to this: (1) it can 
cause some resentment on the part of the salesmen if they feel they are losing 
credit for the sales in question, although this can usually be minimized by achiev- 
ing understanding in advance; and, much more serious, (2) it can distract from 
the field manager's more important, essential management role. 

The fact remains that the economics of the situation may require that man- 
agers do some selling. It may be, for example, that there is not enough potential 
business to require more than a few salesmen in a territory and, therefore, to 
justify the full time of a field manager as a manager exclusively. In such cases, 
then, to have him do some selling on his own is a necessary compromise- -but it 
must be kept in mind that it is a compromise. It may also be the case, for exam- 
ple, that a choice must be made between having close supervision part time and 
distant supervision full time; here the former may be the desirable alternative. 

The question often arises whether in a district or branch, where there are 
really two separate sales forces (being differentiated by product or company divi- 
sion), there should also be two different managers. The answer seems simple, al- 
though, unfortunately, not very easy: yes, if at all possible; yes, even if it means 
some waste of management time, although, obviously, at some point the waste is 
too great to afford. The relevant principle is that the investment in the sales force 
is usually greater than that in field management, and that it is worthwhile to 
capitalize on the greater investment by direct and strong leadership. And the sales 
manager, split between two loyalties, is less likely to be a strong leader. 

The fact is that there actually may be more for the manager to do than is 
realized; in which event, questions of taking time for selling or split loyalty vanish 
into thin air. He should be busy working with his salesmen. Many salesmen are, 
by nature, nonplanners. Yet studies show that, to be more effective, they must 
do more planning, this being the one factor that most separates successful sales- 
men from unsuccessful or mediocre salesmen. There are two very good reasons: 

(1) The typical salesman can actually organize his calls and his travel 
to get anywhere from 25 per cent to 50 per cent more use out of the same 
time by a combination of planning the most efficient schedule and having 
alternate next-best schedules ready when the first one does not work out. 
(Most salesmen do not cut short their losses when they get delayed by having 
to wait outside the customer's office.) 

(2) The most effective sales presentation is one that is thought through 
in advance, so as to get the most cogency into a minimum of words and, even 
more important, so as to be tailored to the customer's specific problem or 
personality. In particular, just because no call can be planned rigidly in ad- 
vance, it is necessary for the salesman to anticipate objectives and be prepared 
to turn them to advantage instead of becoming nonplused, apologetic, or 
even angry. In view of all this, it is amazing how many salesmen go in to 
make a call without more than a general idea of what they are going to say. 



EXPANDING DEMAND 267 

Here is one of the manager's greatest responsibilities, greatest opportunities, 
and greatest challenges. Here is where initial training and continual guidance be- 
come one and the same thing. Here is why, in so many companies, the field sales 
manager is either the weak link or the strong link in the whole selling chain. 

From time to time, he travels periodically with his salesmen; he shows how it 
can be done, by his own example; he asks the salesmen what they expect to say 
before going in, what they think of the results after coming out. Periodically, he 
calls for analyses of competitors' strengths, for potentials of next year's business; 
periodically, also, he measures their performance against expectations (usually 
based on territory potential as described on page 255, modified by his appraisal of 
particular local difficulties and local opportunities) and reviews their records with 
them individually; and so on. Day in and day out, he works and lives with his 
men and has the deep satisfaction of seeing them grow, both to their own great 
good and the company's. 



CHANNELS OF DISTRIBUTION 

The problems of securing effective selling action are compounded when the 
manufacturer does not sell directly to the users or ultimate buyers of his product 
but deals through wholesalers, retailers, and/or other intermediaries. 

It is helpful to think of products flowing from their manufacturing source, 
down through various channels of trade (made up of different combinations of 
different kinds and numbers of intermediaries), to their multiple destinations in 
the hands of individual buyers. Marketing decisions in this area are crucial for 
two major reasons: 

(1) The number and degree of interest of the intermediate distributors 
and final outlets will determine how well they handle the product— carry it 
in inventory, give it sales push, and service the buyers. 

(2) The kind of dealers, wholesale and retail alike, will have a strong 
bearing on how the product will be regarded by buyers on down the channels 
of trade to the market, and this will of course affect the speed and strength 
of the distribution process. 

Involved are some of the same types of problems as discussed in connection 
with salesmen. The distributors and retailers must be chosen to fit the needs and 
opportunities of the product-market situation. Thus, with a highly technical 
product, a distributor (agent, wholesaler, or whatever) must be such as to instill 
confidence in his ability to help his customers with complex problems; he must 
have a high degree of engineering orientation, even at the sacrifice of some sales 
drive. On the other hand, the distributor selling a uniform product like sugar 
to candy makers, bottlers, ice cream manufacturers, and so on, will do better if he 
is the kind that emphasizes careful and conscientious service. 



268 



EXPANDING DEMAND 



There are also a number of additional problems here, such as how direct to 
sell (that is, whether or not to by-pass intermediaries) and what strategy to follow 
at the various stages. 



Role of Intermediaries 

Except for agents or distributors who handle nothing but his product, no one 
else but himself will have the same degree of interest in pushing his product or 
will have salesmen with specialized background for that particular product. But 
usually they can offer him a wide range of customers, whom they can afford to 
contact frequently, since they have a chance to sell not only his product but many 
other products as well. 

In recent years there has been a trend for more manufacturers to bypass whole- 
salers and sell direct to retailers. (Few manufacturers have attempted to go the 
whole way and bypass the retailer, too. There are some notable examples of suc- 
cessful direct-to-consumer selling in such fields as apparel, housewares and house- 
hold equipment, cosmetics, and specialty foods, but altogether they do not account 
for more than one per cent of all goods sold at retail. 24 ) 

This is not because the manufacturer is seeking to eliminate the "waste" of 
an unnecessary function. The need for salesmen to call on retailers (or, in the case 
of industrial goods, users) and for warehousing, credit, delivery, and so on, remains. 
What has happened is simply that the manufacturer has decided that it would 
be a good risk to attempt to do the wholesaling job himself. In some cases he will 
be successful in reducing the expense of operating the necessary wholesale activity 
and/or in increasing volume enough so that the cost per item handled is lower; 
in others he will not. 

As Professor Edwin H. Lewis of the University of Minnesota has pointed out, 
the attitude of manufacturers represents a basic competitive drive: the desire to 
maximize profits by performing any or every function more efficiently. Thus, many 
manufacturers have found it expedient to undertake the wholesale function them- 
selves in circumstances like the following: 

1. Where the product has a high degree of style element or physical 
perishability. (In such situations, the closer the manufacturer is to the market 
timewise or in number of marketing steps, the less his risk is likely to be.) 

2. Where it is necessary to design or modify the product to meet the 
customer's needs, or where the customer expects a considerable amount of 
servicing. 

3. Where the wholesaler is not anxious to push the product because it 
is not important to him or may even be in competition with a private brand 
of his own. 



24 See Victor P. Buell, "Door-to-Door Selling," Harvard Business Review, May-June 1954, 
p. 113. 



EXPANDING DEMAND 269 

4. Where the margin on the product is enough to support the cost of 

more concentrated or specialized selling. 

Over the years wholesalers have had to become content with a smaller and 
smaller profit margin (in most cases counting on increased volume to hold up 
their total dollar profits); see Table 4-C. This is especially true for "con- 
venience goods"— drugs, dry goods, groceries, hardware, and tobacco products— 
which are stocked in as many outlets as possible for the convenience of the con- 
sumer (who buys these goods wherever it is convenient to do so), and which 
represent the particular fields where the wholesaler has always occupied a 
dominant position. 

It is obviously difficult (and in many cases impossible) for the manufacturer 
to perform the basic wholesale function more economically than it can be done 
by independent wholesalers operating at such profit rates. Indeed, if the manu- 
facturer bypasses the wholesalers, it is likely to be for the higher profits that come 
from more aggressive selling, better control of distribution to capitalize on con- 
sumer advertising, and so on, rather than for lower operating costs. But this 
hardly lessens the pressure on wholesalers to step up their efficiency, particularly 
since they are also in a squeeze from large-scale retailers on the other side. 

Chains (both the corporate type like A & P and the voluntary type where 



TABLE 4-C. RATIOS OF NET PROFITS TO NET SALES 


OF 


WHOLESALERS 








(Median figures) 


i 






Kind of Business 


1947 


1950 


1955 


1960 


Auto parts and accessories 


4.41% 


3.94% 


1.71% 


1.74% 


Tobacco products 


0.99 


0.56 


0.36 


0.68 


Confectionery 


2.72 


0.86 


1.53 


1.02 


Drugs and sundries 


2.38 


2.21 


1.70 


1.26 


Dry goods 


2.95 


2.94 


1.00 


0.84 


Electrical parts and supplies 


3.89 


2.97 


1.50 


0.96 


Fresh fruits and vegetables 


1.64 


1.01 


0.94 


1.21 


Men's furnishings 


2.28 


4.88 


2.70 


0.80 


Petroleum products 


3.99 


3.50 


1.29 


2.67 


Groceries 


1.47 


1.39 


0.69 


0.59 


Hardware 


3.11 


2.81 


1.81 


0.99 


Hosiery and underwear 


4.13 


2.59 


1.59 


0.14 


Lumber 


3.25 


2.68 


1.07 


0.92 


Lumber and building material 


4.46 


3.15 


1.54 


0.84 


Meat and poultry 


2.96 


1.42 


0.94 


0.75 


Paints and varnish 


4.08 


0.68 


1.70 


2.58 


Paper 


3.54 


2.85 


1.12 


1.20 


Plumbing and heating supplies 


4.79 


3.68 


1.88 


1.42 


Shoes 


2.01 


2.12 


1.01 


1.35 


Wines and liquors 


2.66 


1.45 


0.86 


1.45 



NOTE: Net profits are profits after full depreciation, inventory and bad- 
debt adjustments, and after all reserves including federal income taxes, but 
before dividends and owner withrawals. Net sales are sales after deductions 
for returns, allowances, and discounts from gross sales. 

SOURCE: 1947 and 1950, data from Dun and Bradstreet, Inc., Behind the 
Scenes of Business, Revised Edition, 1952; 1955 and 1960, data from Dun's 
Review and Modern Industry, November 1956 and November 1961. 



270 EXPANDING DEMAND 

independent retailers band together), consumer cooperatives, mail-order houses, 
and large department stores, because of the types of goods purchased and the 
quantities involved, usually find it advantageous to deal directly with manu- 
facturers. They have no need to buy from wholesalers, except for occasional fill-in 
orders or merchandise controlled by wholesale franchises. In this competitive 
situation, all the wholesale merchant can do is to try to help the retailers who 
buy from him to hold their position— not that this is insignificant, for many 
wholesalers have substantially improved their price and merchandise policies to 
this end. 

Private distributor brands, sold below the price of nationally advertised manu- 
facturer brands, have been widely used by wholesalers to stimulate sales and help 
their retailers meet competition. This is particularly true in the grocery field, where 
distributor brands have been an important part of voluntary-chain operations. 
When they are of high quality, fair in price (which usually means lower in price 
than comparable national brands), and properly promoted, they usually receive 
good acceptance among consumers; and once they are established, they help to 
bring customers back to the particular stores that carry them. On the other hand, 
one of the strong appeals of a store is to offer a wide range of nationally adver- 
tised brands at competitive prices; so there is a limit to which private brands can 
be carried, and the retailer usually considers margin, sales volume, and the impor- 
tance of quality in making his decision on whether to add a private brand to any 
of his lines. 

Although the share of the market handled by manufacturers' branches has 
increased over the years, both the number of wholesale merchants and their dollar 
sales have increased substantially. This would seem to indicate that wholesalers 
have proved to perform a useful economic service, in many instances. 25 

Strategy of Distribution 

Whether the manufacturer uses (1) brokers or agents (who do not stock or 
take title to the merchandise but receive a commission from the manufacturer for 
arranging the sale, and who usually represent a number of noncompeting lines), 
or (2) various forms of wholesale distributors, or (3) various forms of retail stores 
or dealers, or all three, there are a number of things he can do to get more drive 
in the selling efforts in behalf of his products on down the line. For example: 

The manufacturer can provide assistance in the form of displays and 
merchandising aids; send out his own "missionary" salesmen to call on im- 
portant prospects with the broker's, wholesaler's, or retailer's salesmen; arrange 
contests or incentive plans for those who handle or sell his product; and so on. 

He can offer his product on an exclusive basis to selected brokers, whole- 



25 For the full study, from which this material is drawn, see Edwin H. Lewis, "Comeback 
of the Wholesaler," Harvard Business Review, November-December 1955, p. 155. 



EXPANDING DEMAND 271 

salers, or retailers. Thus, he sacrifices some coverage of the market for the 
sake of the extra push that those who represent him exclusively will exert for 
his product because they are assured of all the business in their territories. 
(Only some aren't interested in operating on this basis because they don't 
want to obligate themselves to push this or that product.) 

And he can provide larger margins than are usually made available, in 
the hope that this will stimulate more enthusiasm in behalf of his product. 

Sometimes, two or all three of these approaches are taken, since they reinforce 
each other. Much will depend, here, not just on the strategy the manufacturer 
fancies but, once again, on the product and its market. Some products need less 
push, either because they are already established on the market or because they 
answer some existing need; some need more, for the contrary reason. 

Also, there will be differences in the degree to which the product should be 
carried extensively by a large number of wholesalers or retailers, as against a smaller 
number of selected outlets. If its sales depend on its being available or getting 
mentioned and displayed at the moment it is needed, then the emphasis will be 
on more outlets. If it is the kind of product people will make a little effort to get 
when they want it, and /or it actually gains prestige from being handled by a few 
particular kinds of outlets, then the opposite will be the case. A good example of 
the latter is the way Corning Glass Company increased the sales of its high- 
priced Steuben glassware (often bought for wedding gifts and other ceremonial 
presentations) by reducing the number of stores where it was sold to only a few. 

But whatever the manufacturer does, he is considering the cost and effective- 
ness of these measures for getting more push in back of his product as it moves 
down the line to the ultimate buyers— and not just against alternative forms of push, 
but against the cost and effectiveness of measures to pull through the channels of 
trade (that is, advertising that will lead the buyer to want it, so the dealer will 
want it, so the distributor will want it, and so on up the line) . 

ADVERTISING 

Advertising is one of the prime ways by which American business has ex- 
panded markets. The question of how much to spend on advertising is, however, 
still one of the fuzzy areas of marketing decision making. The biggest problem is 
in tracing the connection between advertising and sales. This does not apply, of 
course, to direct-selling advertising ("send five dollars and the coupon to get your 
first book," and so forth) or to promotional advertising by retail stores. Such 
advertising not only can be evaluated by dollar results, but it usually has to pay 
its own way in immediate sales. But elsewhere it is very difficult, if not impossible, 
to relate advertising and sales dollars. 

For one thing, the effect of advertising is indirect or delayed; it seldom triggers 
the buying act but simply moves people one step closer to purchasing (see Case 



272 



EXPANDING DEMAND 



4-6) or makes them more susceptible to other, more direct influences like display 
or a salesman's presentation. For another thing, its effect is intangible, it is in 
people's minds, and gradations or accretions in attitude are difficult to identify 
and measure. Then, in part, because of these same indirect, intangible qualities, 
any individual advertising effort adds in with other advertising efforts, and, indeed, 
with everything else that the company does in marketing, to produce a growing, 
cumulative impact. In a very real sense, it has to be judged like a long-term in- 
vestment that will pay for itself gradually over a span of time reaching far into 
the future. 

John Wanamaker, the great retail merchant of the last century who pioneered 
in the use of advertising and such other retailing innovations as the one-price 
policy and "all goods guaranteed," is supposed to have said: "I know that half 
the money I spend on advertising is wasted, but I don't know which half." 

The hope is that someday a combination of communications research and 
operations research models, with the aid of advanced, high-speed computers, will 
be able to bring more light on such questions; and much work is now being done 
along this line because of the obvious importance, in terms of dollars, to business. 
Total advertising volume is estimated by Printers Ink, in its "Guide to Marketing 
for 1962," at around 12 billion dollars per year (a little over 2 per cent of the 
Gross National Product). 

In the meantime, there is much that the marketer can do to get some gen- 
eral magnitudes and, perhaps even more important, to make sure that he gets as 
much value as he can from what he does decide to spend. Thus, he can bring 
both marketing research and his own logical analysis to bear on these questions: 

( 1 ) What is the best way for me to spend my dollars? More specifically, how 
much should I put into advertising (or other forms of promotions) as contrasted 
with other ways of influencing prospective customers? 

Thus, in Table 4-D, a manufacturer selling through retailers can compare 
(in the light of estimates of sales, whether intuitive or backed up by marketing 

TABLE 4-D. PROFIT ANALYSIS OF DIFFERENT MARKETING MOVES 



Manu- 
facturing 
Cost per Unit 

1. .50 

2. .50 

3. .50 

4. .50 

5. .45$ 



B 

Price 
to Retailers 

(Return to 
Manufacturer 

per Unit) 

1.20 

.90 

.60 
1.20 

.60 



Retailers' 
Price 

1.89 
1.79** 

.99 
1.89 

.99 



Promotional 
Outlay by 

Manu- 
facturer 

$ 50,000 

50,000 

50,000 

250,000 

250,000 



Fixed 
Cost of 

Manu- 
facturer 

$ 50,000 
50,000 

150,0001 
50,000 

250,000t 



Estimated Profit to 
Sales in Manu- 

Units facturer 



1,000,000 
2,000,000 

10,000,000 
2,000,000, 

20,000,000 



5 600,000 

700,000 

800,000 

1,100,000 

2,500,000 



*F (B-A) - (D + E) 

♦♦Retailers pass on a little of the larger margin to their customers. 

t" Fixed costs" (rent, supervision, heat, light, etc), higher, but not proportionately so, at substantially 
higher volumes. 

:}: Reflecting production economy at this volume. 



EXPANDING DEMAND 273 

research) the different profit effects of a wider margin to retailers (Row 2), a 
substantially lower price to consumers (Row 3), increased promotional outlays 
for advertising (Row 4), or lower price and promotion combined (Row 5). 

(2) Having decided to make use of advertising (or other promotion), what 
exactly do I expect it to do? 

The marketer needs to know what kind of people he will be trying to in- 
fluence, what kinds of appeal will be effective with them, and through what media 
he can bring those appeals to their attention. This question must also be con- 
sidered subsequently by the advertising agency or whoever else prepares the copy 
and the details of the schedule of insertions or programs, but the marketing strate- 
gist must also ponder them, and work out at least rough answers, because the 
whole question of how much advertising to do depends in large part (as we shall 
see) on how much influencing is possible and on whether satisfactory media are 
available. 

The question of how specific and specialized the appeal is has considerable 
bearing here. The more closely it is matched to people's interests and predisposi- 
tions, the more likely its transmission is to survive the four resistance blocks 
discussed earlier. Also, because many people's predispositions reflect their group 
norms, and because, in turn, group norms are affected by the individuals who are 
"opinion leaders" in their groups, and because, finally, opinion leaders are more 
likely to expose themselves and be receptive to the media addressed to this par- 
ticular group, there is additional reason why messages appealing to specific kinds 
of people are likely to be more influential. 

Also of interest in this connection is the finding of Katz and Lazarsfeld that, 
as far as its effect in marketing is concerned, social status is not, as often supposed, 
a strong determinant of opinion leadership, compared with certain other determi- 
nants like stage in the life cycle (married women with families, for example) or 
degree of gregariousness (active members of organizations, for example). That is, 
in any given social stratum, opinion leadership is likely to come from certain 
kinds of people within that particular stratum. 26 Points like this, of course, have a 
bearing on the question of what media to use, as well as on how much the market 
for any product is subject to being influenced. 

(3) In particular, how much of the influencing job do I expect the adver- 
tising to do? This is the question of whether to build a feeling of dependability 
or prestige among buyers; or to get inquiries, that is, names of prospects for sales- 
men to follow up, and perhaps make sure there is a welcome mat out for the 
salesman when he calls; or to activate retailers, as even much advertising in con- 
sumer media is designed to do, rather than to really influence consumers; or what? 
And there is the question of degree: considering the need, in the light of the rest 
of the marketing program, and the cost, in terms of available media and the 



Katz and Lazarsfeld, op. cit., p. 324. 



274 EXPANDING DEMAND 

influenceability of buyers, what is the objective— recognition, acceptance, or actual 
buying urge? 

Perhaps the most rigorous form of question here is: What can be done best 
by advertising, or perhaps only by advertising, out of everything that has to be 
done to move the goods into people's hands? (See page 355 in Section Five for 
a suggestion on how to analyze this question quantitatively.) 

(4) Now, with a clearer idea of what I am trying to accomplish through 
advertising, how much shall I spend? 

Under (2) has been described the so-called "task" approach: What task must 
be performed by advertising? And that basically is what also should determine 
how much to spend on it. But it is possible to go a little further, by systematizing 
the factors that affect the feasibility of performing the task. These factors are of 
two kinds : 

(a) From the advertiser's side, there is the question of how much he can 
afford to spend in terms of the return from the sales effectuated by the advertising 
(see columns i and ii in Table 4-E); and the cost of getting messages through to 
prospective customers depending on whether the people sought are covered eco- 
nomically and effectively by available media (column iii). 

(b) From the buyer's side, there is the question of "how communicable to" 
he is, depending on how much financial stake he has in the purchase (columns 
iv and v); and how important the product is to him psychologically, either because 
he needs to have assurance of quality, having no way of knowing for sure himself, 
or because what the product does for him is a significant part of his way of life 
or business (column vi). 

Consider, for example, how these six factors might apply to a number of 
different products and services, in Table 4-E. There should be no thought that 
Table 4-E is necessarily correct in detail; one could argue about any of the in- 
dividual ratings. A simple plus or minus does not represent enough discrimina- 
tion, either. But the factors are real, and the systematic framework may be helpful. 

To the extent that the payoff of advertising can be estimated in dollar sales, 
the same kind of calculation could be made as for actual salesmen, in Table 4-B. 
Or some idea of the optimum amount to spend on both kinds of promotion could 
be devised by first using a matrix like Figure 4-F; and then finding out, as done 
graphically in Figure 4-G (in the same real company circumstances as Table 4-B), 
what combination of salesmen and advertising would yield the maximum operat- 
ing profit. Note how adding salesmen initially is very effective and then, when 
the important customers are all covered satisfactorily that way, advertising begins 
to add more impact. 

For simplicity, Figure 4-F represents one territory— either a typical territory 
or, if there is much variation, a territory of median size and potential. And, for 
convenience, advertising expenditures are considered in units of $20,000— equiva- 
lent to the cost of a salesman. The cells of the matrix should contain the best 
possible estimates for the combined effect of salesmen and advertising effort. Here, 



EXPANDING DEMAND 



275 







* 






*42 






.*$■£ 






•^ s> 






ft 






^3 


Ho 

5; 




^ 


eo 






0) 






ttj 




gical 

ce of Heavy 

ty Advertising 




s 








ooO 1 


rS 




^■cx 


CO 




ft, 5 


K 






<*> 






>. 




**> 0) 


3 




^ CO 


05 




quen 
of 
rcha 


s 


ft 


I s, 









u 






&H 




<4) 

CO 




ft 








ft, 




^ 


? ft 






Cj -^ 






•^i So 




:j3 « 


SO B 






0*5 


tt 


e 




a 
^ 


*1 


V 






0) 


CO 




0) 
















ddby 

ticular 

ertiser 


c» 


fc 


03 e *« 


►si 


►o 




£ 


ft, ^ 


$ 


s 




o 


£; 




!s 






fc, 







S"S 


























^ 




























' 


1 


i 


4 


+ 


1 


+ 


i 


1 


• 


+ 


+ 


§ <^> 


























^ ft, 




co 






CO 


CO 

cu 

CO 
CO 










CO 

o 

>- 








o 




a. 

-a 


CD 
> 


d 








CO 




C 






OS 




o 


d 


bn 


CO 






^ 


CO 







CO 

u 

a 
h 


a 

o 

CD 

1 I 


a 

a; 

■a 


o 

"cd 
B 

Jh 

3 
O 


43 
CO 

u 

cu 


c 

3 
co 

3 

CO 

a 


^0) 

o 

a 

o 

3 


a 

1 

o 

jG 
V, 


u 

a 

3 


o 
a 

a 

1 


c 

CD 

s 

to 

> 


o 
o 

c 




U 


M 


o 


o 


W 


U 


< 


5 


5 


5 


S 



276 



EXPANDING DEMAND 



the company's existing sales records will probably not yield much help, except 
in the case of the one column showing sales volume resulting from the salesmen's 
effort and no advertising. Either management must rely completely on its "feel" 
for what a given amount of advertising coming into the territory would do in 
combination with salesmen (covering some of the gap left by salesmen, helping 
them to get in to see prospects and to be more effective, or in whatever way); or, 
again, management must employ deliberate experimentation, whereby, through split 
runs, regional editions, local stations, and so forth, the amount of advertising is 
varied in test areas versus control areas and the resulting variation in sales volume 
measured. 



Number of 
Salesmen 



Units of Advertising 
B 




Figure 4-F. Table of Sales Volume Produced by Salesmen and Adver- 
tising in One Territory with Potential of One Million Dollars. 

Note: If salesmen cost $20,000 each and advertising units $20,000 each, then the shaded 
areas denote six combinations that are optimally effective. Figures within cells are sales units 
of $10,000 each. Thus, one salesman can produce up to $250,000 sales; 2 salesmen up to 
$500,000; 2 salesman and 1 unit of advertising up to $600,000; and so on. In the case of 
$900,000 sales volume, where 4-2, 3-3, and 2-4 show as equally effective combinations, 
3-3 has been chosen as being the practical compromise. Other combinations can be dis- 
regarded because they would be more expensive per dollar of sales. 

The experimental approach implies that current advertising has an effect on 
current or at least near-current sales. That is all that would matter with intro- 
duction of new products, for instance. In other cases, the findings of tests (unless 
continued over a considerable period) would still have to be extrapolated for their 
long-run effect. 

Particularly because of the delayed effect of advertising, and, over time, the 
cumulative impact of many delayed (and continuing) effects, management must 
make the necessary adjustments in its estimating. It might be possible to conceive 
of advertising as having already been in force for some time and producing an 
effect like the interest on a capital investment made in the past. Or, in the event 
that management must make a new decision about advertising, the proper ap- 
proach might be roughly similar to the discounted cash flow method of analyzing 
a financial investment. Or, for a contemplated advertising budget of $100,000 



EXPANDING DEMAND 



277 



$7 



in 




rr 




_i6- 




_i 




o 




o 




u. 




o 


- \ U / 


</> 




2 




O 




ds- 


"* 


s 




z 




^ 




z 
o 


I S 


1- 




04- 


- X 


2 







- 


Q. 




£T 


" n l 


O 




U. 




to 




rr ° 




^ 


d 


1- 


■ V 







z 


- \ e / 


UJ 




Q- 




X 




UJ 




0<* J 




Z 




< 




1- 




u_ 




O 




(T 




Q. 




O 1- 




Z 




h- 




< 




or 




UJ 




CL 




O 





'1-6 



KEY 

BASED ON SALESMEN PER TERRITORY 
IN FIGURE 4-F x 25 x $20,000 FOR 
OPTIMUM COMBINATIONS 

BASED ON ADVERTISING UNITS 
PER TERRITORY IN FIGURE 
4-F X 25 x $20,000 FOR 
OPTIMUM COMBINATIONS 

A-HB 

EXPENDITURES FOR SALESMEN 

EXPENDITURES FOR ADVERTISING 

SIX OPTIMUM COMBINATIONS FOR 

DIFFERENT SALES VOLUMES, 

FROM SHADED CELLS 

OF FIGURE 4-F ^K' 

SEE NOTE ON TABLE 4-B <&? 

<£/ 

SALES VOLUME FOR Q> 

TERRITORY IN AC/ 

FIGURE 4-F x 25 rVV 

/ 



A. 






■<?? 



sv 



# 



& 

W 



MAXIMUM DOLLAR PROFIT 
WITH APPROXIMATELY 75 
SALESMEN AND $100,000 
ADVERTISING EXPENDITURE 
(SECURING 90 7o OF THE COMPANY'S, 
POTENTIAL MARKET. ) 



DOLLAR PROFITS AFTER 
PROMOTIONAL COSTS 



CO 



&% 



*& 



Xr 



$£^0***® 



fOft 




< — S-— -x— H-- 
4 5 




8 9 10 II 12 13 14 15 16 17 I8 Q 19 20 21 22 23 24 25 
SALES ( IN MILLIONS OF DOLLARS)^ 



Figure 4-G. Determination of Operating Profit with Varying Number of Salesmen and Size of 
Advertising Expenditures, for All 25 Territories with a Total Potential of $25 Million 



per year, where the annual impact might be expected to level out ultimately at 
$400,000 of sales per year, the development of the selling effect might go like this : 

J si year 2nd year 3rd year 4th year Sth year 6th year 

$200,000 $300,000 $340,000 $360,000 $368,000 $372,000 

This assumes that about 50 per cent of advertising has current effect, plus a diminishing 
effect carried forward from previous years at the rate of 25 per cent from the year before, 
10 per cent from the year before that, and 5 per cent, 2 per cent, 1 per cent, and so forth, 
from previous years. Thus, the sixth-year figure is composed of (.50 + -25 + -10 + -05 
+ .02 + .01) X $400,000. 



278 EXPANDING DEMAND 

In any event, the fact that advertising has so many intangibles and imponder- 
ables in it makes it all the more important to try to develop some quantitative 
thinking about it— not so much as a dependable basis for making decisions with- 
out other guides (indeed, quite the opposite when the estimates are thin), but 
rather to stimulate judgment, and also to raise some red flags against taking 
habitual or intuitive courses of action for granted. In this sense, any estimates that 
get one closer to a good decision are better than none. 

More elaborate methods would of course be needed for more complex de- 
cisions, such as development of a complete strategy with all elements of the 
marketing mix included. If such methods are available in a form where they are 
enough more powerful to cope with the additional complexity, fine. But otherwise 
it may be better to stick to just a few key elements, where both estimates and 
judgment are more clear-cut and reliable, and be content with quantifying only 
that far. 



Management Responsibility 

Above all, it needs to be emphasized that questions of marketing strategy 
cannot, or should not, be abdicated to specialists in advertising agencies (or to 
marketing research agencies, either). This does not mean that managements 
should not take advantage of the wisdom and experience agency people have to 
offer, but rather that they should not abandon their own responsibility for think- 
ing analytically and creatively. Advertising also does require specialized skill in 
art and language, in conceiving original and striking ways of making points; and 
it is for this that advertising agencies customarily are paid 15 per cent of space 
and time costs, and are worth it. 

Usually, the media bill the agencies at published rates, less 15 per cent, and 
the agencies bill clients at the full rates. Most agencies also charge their clients a 
fee for research or other efforts in their behalf beyond the usual preparation of 
copy, art, etc. Recently, there has been considerable argument against the 15 per 
cent arrangement, on the grounds that it is not enough in some situations, and 
too much in others; or that, because it is a percentage rather than a fixed amount, 
it tends to encourage an agency, consciously or unconsciously, to recommend 
larger budgets and more expensive media. Some agencies have abandoned the 
percentage arrangement in favor of a contractual fee, for all or some of their 
clients, but there is no pronounced trend in this direction as yet. 

In any event, whatever the merit or lack of merit in such arguments, so long 
as management is alert, and plays its role in policy and strategy decisions, there 
is not much reason to worry about being pushed into unwise expenditures. To 
the uninitiated, advertising may seem mysterious and abstruse; as we have seen, 
it is indeed intangible and hard to measure. But, by the same token, it needs all the 
more to be looked at realistically, systematically, and logically by marketing 
management within the company— and, if it is, then the agency, too, can do a 



EXPANDING DEMAND 279 

better job of serving the company in the special areas of its competence and 
experience. 

For the Sake of Selling 

In this book we shall not delve into the intricacies of advertising— into ques- 
tions of frequency; duplication; color versus black and white; spread, or full-page, 
or part-page; coverage versus penetration; and so on. (We shall, however, briefly 
describe in Section Five how measurements can be applied to this area.) But it 
should be worthwhile to emphasize here that management (and perhaps special- 
ized advertising people, too) can profit from thinking of their pieces of advertisng 
as salesmen— no matter how much or how little of the full selling job they are 
supposed to do. 

Thinking of pieces of advertising as salesmen is a good safeguard against be- 
coming preoccupied with "coverage" and "impact" and the other criteria of ad- 
vertising as ends in themselves. "Coverage" means nothing unless it represents 
the kind of people that the advertiser wants his salesmen to call on— and if he has 
a limited budget, he should direct them first to his prime prospects. "Impact" 
means nothing unless it signifies that the salesman has been admitted to make a 
presentation— and if, for example, he makes a pleasant impression but exerts no 
selling influence, he might as well have stayed home. (In the case of consumer 
goods, just the presence and display of the product in a store is also like a sales- 
man—another salesman, helping the advertising.) 

And don't forget that the same salesmen in different media may act differ- 
ently, too, and get to people at different times and in different moods. Thus: 

The newspaper-ad "salesman" is welcomed by housewives about to go 
shopping and looking for news of department store items; that is why many 
manufacturers have cooperative arrangements with retailers, whereby the manu- 
facturer pays a portion (usually 50 per cent) of the retailer's expenditures in 
behalf of the product, thus encouraging the retailer to advertise (at the local 
retail rate, which is less than the rates to national advertisers) and doubling 
(or more than doubling) the space that he gets for every dollar of his own. 
As for the television-commercial "salesman," he is welcomed into the 
home when both the wife and husband have finished work and is especially 
good at presenting products that can be demonstrated visually, like an auto- 
mobile wax. On the other hand, the magazine-ad "salesman" may stay for a 
longer time and get several chances to make his points, which is particularly 
useful on something requiring considerable deliberation, like a once-a-lifetime 
tour of the world. 

Then, also, within each group of media, the "salesman" sent out through 
one magazine, or one TV network, or one newspaper, may get a better 
hearing than the salesman from another, because the medium that brings 
him commands more respect or elicits more attention for that particular 
product. 



280 EXPANDING DEMAND 

A salesman in the flesh has the advantage over a salesman made of paper 
and ink or of electronic pulses. He can, if he is adept, keep after his prospect and 
fashion his presentation to fit the needs of the situation as it develops. But the 
advertisement-salesman has only one quick chance to "get his licks in." If he doesn't 
succeed then in making some selling impression, however slight, that call is wasted. 
As a matter of fact, his employers don't really expect him to be successful on 
more than a small percentage of his calls— and they judge him on whether he can 
gradually add to that percentage. 

Nevertheless, the same reasons that some people have for buying a product 
are the reasons that lead other people to even look at the product, then to begin 
to get interested, and, eventually, to contemplate buying. So, just because adver- 
tisement-salesmen do not have the opportunity, or the discipline, of being able to 
tell how well they are doing as they make their calls, it behooves their employers, 
that is, marketing management, to see to it that, when they go out on their calls, 
they are as well-prepared salesmen as they possibly can be. 

In short, the end objective is sales, and the only way to get sales is through 
buying action, or through some change in how people feel about a product so 
that they move that much nearer buying action. The advertising message (does 
it cause people to change their desires or attitudes?), the people it is addressed to 
(are these the ones who implement change by making or influencing buying de- 
cisions in their company or social group or family?), the media that carry it (will 
they bring the message to the right people, at a time when they are in the frame 
of mind for thinking about the kind of change this product represents? ) —all will 
be more effective if they are judged, like salesmen, for their ability to bring about 
or facilitate change in people. 

The efforts of advertising may be less direct and less tangible, but they too 
are more likely to pay off in attention, interest, and actual purchase — more likely 
to get through and to make a definite impression— if they are focussed on people's 
needs and wants as all good selling attempts must be. For it is needs and wants 
that form the basis for change; in fact, change is another way of saying goal. 

There is evidence, too, that when people are ready or have to make a change, 
they rely on advertising for guidance. A study of what brands people choose when 
their favorite brand is not available reveals that almost invariably they turn to 
the most heavily advertised brand, no matter what their previous loyalty. This is 
a tribute to the power of advertising; it indicates that, when other predispositions 
do not interfere, the weight of advertising is itself a basis of predisposition. 27 



27 See Paul M. Carrick, Jr., "Why Continued Advertising is Necessary: A New Explana- 
tion," Journal of Marketing, April 1959, p. 386. 



EXPANDING DEMAND 



281 



CASE 4-1 

Buying an Automobile 



(A man and wife of about 50 
years of age come into medium- 
price automobile salesroom.) 

Man: Saw that convertible in the 

window. . . . 

Salesman: Yes, isn't that a beauty! It 
really catches people's eyes as 
they go past. 

Man: Of course, we wouldn't want 

anything quite that flashy; I'd 
feel a little foolish. But it does 
remind me of a car I had once; 
only then. . . . 

Wife: A little too daring for us. That's 
more the kind of thing that 
our 17-year-old boy would want. 
He likes the sun and the air, 
and, well. . . . 

Salesman: I know how you feel. Inappro- 
priate. Now, take a look at this 
conservative 2-door. . . . 

Wife: But I don't like that uphol- 
stery. It's so drab. 

Salesman: Madam, this is the finest ny- 
lon-plastic, guaranteed to out- 
wear iron; all the cars are 
coming with it this year. Why, 
only yesterday, Mrs. Jones came 
in and. . . . 

Man: What would my trade-in be on 

this one — more than on the 
convertible, I guess, not to 
mention the wear and tear 
every time you put that top up 
and down. 

Salesman: Don't worry, we give you what 
your old car is worth, no matter 
what you buy. Actually, this 
one would run you a little less, 
net. This is a real buy for de- 



pendable transportation. . . . 
(The salesman goes on talking 
about the ruggedness and horse- 
power of the car; economi- 
cal gas mileage; future trade-in 
value, and so forth. . . .) 
[If this is all of the picture 
you had, would you expect the 
man and wife to buy?] 

Man: Of course, with my present car 

I can still. . . . 

Salesman: Let me call the appraiser, and 
while he's looking at your car — 
it's a 1957 model, isn't it? 

Man: No, a 1958. 

Salesman: Well, those old ones all look 
a little alike, I think, particu- 
larly when compared with the 
1961. See, how it. . . . 

Wife: John, I think we'd better be 
getting along. I've got to get 
dinner started. 

Man: Haven't you got any of the new 

economical compacts? 

Salesman: Why yes, let me show you. . . . 

Man: No, don't bother now. We 

can't really afford a new car 
this year. We'll be back again 
later. 



(Next week or so, same man 
and woman are in another sales- 
room — this time, foreign cars.) 

Man: What do you call that little 

midget, in the window? 

Salesman: That's a Bentlinger Bantam, 
so called because, though it's 
little, it has a lot of get-up-and- 



282 



EXPANDING DEMAND 



go. . . . Are you interested in a 
small car, or a foreign car? 

Wife: Well, not particularly, but it 
does look a little unusual, 
doesn't it? 

Man: And I guess it's economical too. 

Salesman: Yes, it is economical on gas, 
and oil, but I think the most 
attractive feature is, well, the 
fact that it looks distinguished, 
different. . . . 

Wife: Our 17-year-old son is more in- 
terested in the sporty kind. 

Salesman: My, you have a son that old? 
I'll bet he does a lot of dating, 
doesn't he? 

Man: Too much, much too much, 

ought to do more studying. 

Wife: Now, John. 

Salesman: I think I know how you both 
feel, though my problem — and 
pride — is a 15-year-old girl. And 
I must admit she does like the 
boys who come for her in a 
nice-looking car. As a matter of 
fact, I've learned a lot about 
cars just from listening to how 
she feels. 

Wife: Does she like these little for- 
eign cars? 

Salesman: Oh, yes. She says most of the 
boys have either the conven- 
tional, unimaginative cars their 
families own, or one of those 
beaten-down and souped-up old 
wrecks — and it makes her feel 
grown-up and important to be 
called for in something like 
one of these Bentlingers. 

Man: You did say it's economical? 

Salesman: Oh yes, and not just gas and 
oil. . . . 

(The salesman goes on to talk 
about maintenance, repairs.) 
[Again, what do you think — 
will they buy?] 



Man: But you don't have any model 

on which you can put the top 
down, do you? 

Salesman: No, the Bentlinger people have 
found that that adds too much 
expense to the initial price and 
the maintenance, but they give 
you the same values with this 
roof panel; and don't forget 
how easy this is to park, par- 
ticularly for the ladies. 

Wife: But I don't drive; John, and of 
course Johnnie, take care of all 
that. 

Salesman: But you do like your men folk 
to be happy and proud, don't 
you? 
Let's see what the trade-in is 



Man: 
Salesman 



on my car. 

My, you've taken good care of 
your car, haven't you? I can 
see you're the kind of person 
who takes pride in a good car. 
Now, you understand, don't 
you, that I can't give you the 
big allowance you might get 
toward a big, conventional car 
— we don't price up our cars 
artificially just so we can im- 
press you with a big trade- 
in. .. . 



Now, do you think they will buy? Did 
the couple really want to buy a car? What 
kind? What happened in the two sales 
presentations? What good or bad devel- 
opments took place and at what points? 
What desirable or undesirable qualities did 
the two salesmen demonstrate? If you had 
to pick one personal quality for successful 
selling in the kind of situation displayed 
above, what would it be? 



EXPANDING DEMAND 



283 



CASE 4-2 

Continental Appliance Company 



In the summer of 1953, the Continental 
Appliance Company retained the market- 
ing consulting firm of Marshall and Com- 
pany to ascertain the sales possibilities of 
three new mechanical dishwashers. 

The Continental Appliance Company 
had under development at the time sev- 
eral models of dishwashers incorporating 
feaures which it hoped would overcome 
dishwasher sales resistance. Information 
previously gathered had indicated that the 
conventional floor-counter type did not 
satisfy the needs of many American house- 
holds. Among the new types under devel- 
opment were: a roll-around portable, a 
counter portable, and a wall-hung cabinet 
model. 

The roll-around portable was about 
three feet in height and mounted on 
casters. It was filled with water and then 
emptied by means of two flexible rubber 
hoses: one to be attached when needed to 
the kitchen sink faucet and the other to 
rest in the sink basin. The unit could thus 
be moved out of the way when not in use 
and stored in some unused space in the 
kitchen or other part of the house. In ad- 
dition, because it required no permanent 
plumbing or installation, it was felt that 
the roll-around would better serve the 
needs of the apartment dweller or other 
individuals who did not want to invest in 
plumbing when they were apt to move 
periodically. Moving the roll-around from 
home to home would be a simple task 
whereas dismantling and reinstalling a per- 
manent dishwasher is an expensive under- 
taking. Having a capacity for washing 
dishes for six people, the unit was designed 
to sell for approximately $270 as com- 



pared with a current price of $340 for 
the conventional permanent floor-counter 
model. 

The counter portable resembled a large 
bread box and was designed to fit on a 
kitchen counter. It also was operated by 
means of two flexible rubber hoses. It had 
all the portability advantages of the roll- 
around but was designed primarily for the 
small family and the small kitchen lacking 
much free space. When needed, the unit 
could be carried to the counter next to 
the sink basin and afterward stored in any 
free space including a closet in some other 
part of the house. With a capacity for 
washing dishes for only four people, this 
machine would retail for about $150. 

The third new model under devel- 
opment was a wall-cabinet model. This 
model required the installation of per- 
manent plumbing and had to be attached 
to the wall in much the same manner as 
a conventional kitchen cabinet. This unit 
resembled an ordinary overhead kitchen 
cabinet and could be installed over the sink 
or at some other convenient spot in the 
kitchen. Dishes were placed on end in 
racks, the door closed, and the unit set in 
operation. At the conclusion of the wash- 
ing cycle the dishes could be removed and 
placed in a regular dish cabinet or allowed 
to remain in the dishwasher, thus employ- 
ing the unit in place of a regular dish 
cabinet. This model also had the standard 
capacity of washing dishes for six people 
at a price of $250. 

Thus, all three units were envisioned as 
selling under the existing $340 price of the 
conventional floor-counter type, thereby 
possibly appealing to the individual who 



284 



EXPANDING DEMAND 



considered the current price of dishwashers 
to be too high. 



THE STUDY 

After considering the high costs of tool- 
ing and distribution, Continental's man- 
agement decided that a research project 
dealing with public acceptability of these 
new dishwashers would be a worthwhile 
investment before proceeding any further 
with development or actual attempts at 
marketing the units. In addition the com- 
pany wished to obtain some information 
on the dishwasher market in general. 

Planning 

In determining the best means of ob- 
taining the desired information, it was 
recognized that there were three sources 
that had to be approached if a complete 
picture was to be drawn: the final con- 
sumer, the appliance dealer, and the home 
builder. With the growing trend toward 
the sale of new homes with "package" 
kitchens, the latter group was assuming 
greater importance in the appliance mar- 
ket. By speaking with dealers as well as con- 
sumers, a greater "feeling" for the market 
could be obtained since these people 
would be aware of the over-all appliance 
needs of their communities, that is, based 
upon previous dishwasher sales experience, 
whether the new models under considera- 
tion could be sold where the models 
presently stocked could not. 

The research method conceived to pre- 
sent the three machines under study con- 
sisted of the use of a mobile kitchen to be 
driven to the various cities in which inter- 
views were to be conducted. Consequently, 
a demonstration trailer 18 feet long, IVi 
feet wide, and 7 feet high was outfitted at 
one of Continental's plants. When com- 
pleted, it was a replica of a kitchen, com- 



plete with sink, cabinets, and chintz 
curtains. The three new dishwashers were 
then placed in the kitchen: the roll- 
around in a corner where it could be 
moved freely, the counter portable on the 
sink counter, and the wall-hung model in 
place of a conventional cabinet overhead 
to the left of the sink. In addition to these, 
a standard floor-counter model was in- 
stalled below the sink. This was to be used 
as the control mechanism for the question- 
ing on the three new machines. A full 
system was provided for interviewing in 
the evening. 

There were limitations, however, to the 
conditions of reality actually produced. 
The three experimental models were not 
working models. The roll-around and the 
wall-hung models were completely finished 
in white enamel but lacked motors and 
other moving parts, while the counter 
portable was a wooden mock-up painted 
white which also was devoid of moving 
parts. Because of these mechanical limita- 
tions and the difficulty of providing a suit- 
able water supply for the trailer, it was not 
planned to present the dishwashers in 
actual operation. 

The consumer sample chosen was a 
quota based on five characteristics: geo- 
graphic area, community size, home own- 
ership, age of housewife, and economic 
status. A sample size of 800 was settled 
upon as being adequate for the purposes 
of the study. Table 4-H contains the 
breakdown of the sample with regard to 
each of the five characteristics. The sample 
was confined to the top 60 per cent of the 
economic scale since it was felt that this 
group, consisting of the so-called pros- 
perous, upper middle, and middle class 
groups, constituted the market for me- 
chanical dishwashers. 

The dealer study consisted of interviews 
with 100 dealers drawn mainly from lists 
supplied by Continental distributors. The 






EXPANDING DEMAND 



285 



only quotas established for the interview- 
ing team were those of geographic area 
and community size, inasmuch as a cer- 
tain number of dealer interviews were to 
be obtained in each community visited. 
Actually, the interviewing team attempted 
to obtain interviews with the more im- 
portant appliance dealers in each com- 



munity, together with a sprinkling of 
smaller, average-volume merchants. 

The sample for the builder study was 
set up in about the same manner as the 
dealer study with quotas established only 
by city. The sample size was set at 50. 
Here again, the interviewing team at- 
tempted to interview the more important 



TABLE 4-H. CONTINENTAL APPLIANCE COMPANY 
Characteristics of Consumer Sample Characteristics for Which Quotas Were Established 







Number 


Percentage 




Quota 


Interviewed 


Attained 


Geographical Location: 








Eastern States (Mass., N.Y., N.J., Pa.) 


50% 


400 


50.0% 


Central (Ohio, Ind., 111.) 


30 


259 


32.4 


Southern (D.C., Va., N.C.) 


20 


141 


17.6 




100% 


800 


100% 


Community Size: (city population) 








1,000,000 and over 


60% 


482 


60.2% 


250,000 to 1,000,000 


15 


119 


14.9 


100,000 to 250,000 


15 


129 


16.1 


25,000 to 100,000 


10 


70 


8.8 




100% 


800 


100% 


Home Ownership Status: 








Home owners 


60% 


491 


61.4% 


Renters 


40 


309 


38.6 




100% 


800 


100% 


Age of Housewife: 








Under 30 years old 


30% 


236 


29.5% 


30-39 


35 


282 


35.3 


40-49 


17% 


142 


17.8 


50 years and older 


17% 


138 


17.2 


Age undetermined 




2 


.2 




100% 


800 


100% 


Economic Status: 








Prosperous (upper) 


10% 


83 


10.4% 


Upper middle 


30 


236 


29.5 


Middle 


60 


481 


60.1 




100% 


800 


100% 


Other Characteristics 








Type of Dwelling: 








1 family houses 


65% 


511 


63.9% 


2-4 family houses 


10 


99 


12.4 


Apartments 


25 


190 


23.7 




100% 


800 


100% 


Number of People Living in Household: 








1 


- 


10 


1.3% 


2 


20% 


145 


18.1 


3 


30 


224 


28.0 


4 


30 


235 


29.4 


5 


15 


133 


16.6 


6 or more 


5 


53 


6.6 




100% 


800 


100% 



286 



EXPANDING DEMAND 



builders in the community, when possible, 
while not overlooking the smaller builder. 
Most of the interviews were conducted 
with builders of middle- and higher-priced 
homes, as these seemed to be the logical 
market for mechanical dishwashers. 

The initial plans called for the trailer 
unit to visit 19 interviewing areas in 9 
states east of the Mississippi. 

TABLE 4-1. CONTINENTAL APPLIANCE 
COMPANY 

Characteristics of Dealer Sample 



Personnel Interviewed: 

Owner -sales man 45% 

Manager- salesman 39 

Key sales personnel 8 

Non- sales (buyer, owner, manager) . . . 8 

100% 
Type of Outlet: 

Household appliance store 65% 

Plumbing supplies, contracting 14 

Home furnishings store 10 

Department store 6 

Other (hardware, etc.) 5 

100% 
Location of store: 

Central section of city 46% 

Outlying shopping section 54 

100% 
Geographical location: 

Eastern 48% 

Midwest 32 

Southern 20 

100% 
Community Size: 

250,000 and over 68% 

100,000 to 250,000 22 

25,000 to 100,000 10 

100% 
Types of Dishwashers Stocked: 
Stock both permanent and portable . . . 54% 

Only permanent 45 

Sells by order only 1 

100% 
Volume of Sales: (units sold in 1952) 
Very large (60 or more) 19% 

Large (30 to 59) 20 

Medium (10 to 29) 32 

Small (under 10) 29 

100% 



Field Work 

The procedure for performing the field 
work followed approximately the same 
pattern in each city visited. Two staff men 
from Marshall's New York office com- 
prised the basic interviewing team. These 
men were rotated off the project periodi- 
cally, one at a time, and replaced by others 
from the New York office. The two men 
drove a sedan to which was hitched the 
demonstration trailer. Upon arriving in a 
city they telephoned the local female in- 
terviewers who were to accompany the 
Marshall staff men during the next day's 
work. These women had received a de- 
scription of the project and instruction by 
mail prior to the arrival of the unit in 
their town. 

Depending upon the number of inter- 

TABLE 4-J. CONTINENTAL APPLIANCE 
COMPANY 

Characteristics of Builder Sample 

Number 
Geographical Area: 

Eastern 27 

Central 15 

Southern _8 

50 
Type of Builder: 

Speculative building 22 

Speculative and contractual 18 

Contractual building only 10 

50 
Price Class: 

Under $12,000 6 

$12-$15,000 11 

$16-$17,000 9 

$18,000 and over 24 

50 
Types of Homes: 
All types 

Number of Homes Built: 

Under 20 homes 18 

20-49 homes 13 

50 or more homes 19 

50 

Builders installing some dishwashers 74% 
(most in $18,000 and over) 



EXPANDING DEMAND 



287 



TABLE 4-K. CONTINENTAL APPLIANCE COMPANY 
Interviewing Areas 





Consumer 


Dealer 


Bui la 


Boston (Metropolitan) 


X 


X 


X 


Canton 


X 


X 


X 


Charlotte 


X 


X 


X 


Chicago 


X 


X 


X 


Cincinnati 


X 






Evansville 


X 


X 


X 


Long Island 


X 


X 


X 


No. New Jersey 


X 


X 


X 


New York City 


X 


X 


X 


North Adams 




X 


X 


Peoria 


X 


X 


X 


Philadelphia (Metropolitan) 


X 


X 


X 


Pittsburgh 


X 


X 




Pittsfield 


X 






Richmond 


X 


X 


X 


Rocky Mount 


X 


X 


X 


Toledo 


X 


X 


X 


Washington (Metropolitan) 


X 


X 


X 


Westchester County 




X 


X 



views to be obtained and the difficulty 
in securing them, the interviewing day 
would begin at 9 a.m. and last as late as 
11:30 p.m. 

Consumers 

On the days that consumer interviews 
were secured, the two staff men split the 
day, each working half a day with a dif- 
ferent female interviewer. Each female in- 
terviewer was given a final briefing by the 
staff man with whom she was to work, 
which included a complete explanation 
of the purpose of the study, the dish- 
washers under consideration, and the ques- 
tions to be asked. The female interviewers 
were used only on the consumer inter- 
views. 

After arriving on a block, the staff man 
would prepare the trailer while the female 
interviewer began "ringing doorbells" to 
seek out the first respondent to bring out 
to the trailer. Several preliminary ques- 
tions were asked of the housewife in her 
home by the female interviewer. The 



questions included in the consumer ques- 
tionnaire are outlined in the appendix on 
page 291. The preliminary questions dealt 
mainly with the housewife's familiarity 
and experience with automatic dishwashers 
and her current buying interest. The nec- 
essary classification data were also gathered 
at this point. Sections A through D of the 
questionnaire also were covered in this 
part of the interview. The housewife was 
then asked to come out to the trailer, 
where she was introduced to the Marshall 
staff man, who took over while the female 
interviewer left to find the next housewife 
to be interviewed while the staff man con- 
tinued the interview. 

First, the staff man explained the pur- 
pose of the study to the housewife (with- 
out divulging Continental's name). He 
then proceeded with Section E of the 
questionnaire by demonstrating the roll- 
around portable and the counter portable 
and asking her to state a preference with 
her reasons for this preference. The floor- 
counter model and the wall-cabinet model 
were demonstrated next, and the house- 



288 



EXPANDING DEMAND 



wife's preference and reasons elicited. An 
explanation was offered with regard to the 
floor-counter model to the effect that this 
model was to be viewed as merely typical 
of several types currently on the market. 
The main interest was in its position in 
the kitchen (permanently installed un- 
der the sink counter) regardless of whether 
the model happened to be a drop-door, 
top-opening, or pull-out type such as the 
one in the trailer. 

Since the dishwashers carried in the 
trailer were not working models, the dem- 
onstrations mentioned above consisted of 
exhibiting the particular model's porta- 
bility, position in the kitchen, method of 
loading (a full set of dishes and silverware 
were carried in the trailer), and stating 
price and capacity. After establishing sepa- 
rate comparisons between the two por- 
tables and then the two permanently in- 
stalled models, the housewife was asked 
to state a preference among all four 
models and also a second choice, with 
reasons for these preferences. 

Additional questions were then asked 
concerning where the housewife would 
store a portable when not in use (if a 
portable was her first or second choice 
among all four models). Also, if the floor- 
counter model was preferred over the 
wall-cabinet model, she was asked whether 
the preference would change if a dish- 
washer were to be installed in a new 
home. 

Finally, the staff man attempted to get 
some indication as to whether the respond- 
ent's interest in buying a dishwasher in 
the future had changed after seeing the 
new models demonstrated in the trailer. 
After completing several interviews on the 
block, the trailer would be driven to an- 
other block in the same section of town, 
or to another section of town to obtain 
additional interviews. The number of areas 
of a city in which interviews were con- 



ducted varied according to the number of 
interviews in the city quota and the gen- 
eral dispersion of economic groups in the 
city. Usually, an attempt was made to 
visit at least two areas for each of the 
three economic groups for which quotas 
were established. 



Dealers 

The dealer interviews were usually con- 
ducted by both staff men together. The 
trailer would be driven to an appliance 
store picked from the list supplied by the 
Continental distributor or, in many cases, 
to another dealer selected because of his 
reputation in the community as being an 
important factor in the appliance field. In 
the latter category, for example, was an 
interview obtained from one of the na- 
tion's largest discount operators. As a rule, 
the manager or sales manager would be 
sought out for the interview. The purpose 
of the study was explained to him, and 
he was requested to accompany the two 
staff men to the trailer. Most of these 
dealers were anxious to see the new dish- 
washers, and consequently, there was little 
problem of creating interest. 

The initial section of the interview (see 
Appendix 2 on page 292) consisted of dis- 
cussing the present and future sales pos- 
sibilities for automatic dishwashers. Then 
the two portable models were demon- 
strated, and the dealer was asked for his 
opinion concerning the sales possibilities 
of each of these. The same comparison 
was then made with the two permanently 
installed models. Finally, he was asked to 
"guesstimate" how many units of each 
type he thought he could have sold had 
he carried all four during the full year 
1953. 

Throughout the interview one inter- 
viewer would talk with the respondent 
while the other took notes, thus permit- 



EXPANDING DEMAND 



289 



ting a free-flowing discussion devoid of 
interruptions for writing. It should be 
noted that while the questions indicated 
on the questionnaire were covered in de- 
tail during the course of the interview, 
much general discussion was engaaged in 
concerning the dishwasher market, indi- 
vidual problems in selling dishwashers, 
sales methods, and the appliance market 
in general. These comments provided val- 
uable background information for the 
project director in New York. 

Builders 

The builder interviews were conducted 
in about the same manner as the dealer 
interviews. The trailer was driven either 
to the builder's office or to the actual 
building area where he was currently work- 
ing. Portable dishwashers were not dis- 
cussed with the builders, since it was 
assumed that in all cases a builder would 
be interested in making a permanent in- 
stallation of a dishwasher in a new home. 

At the outset of each interview, back- 
ground information was obtained con- 
cerning each builder's attitude toward 
dishwashers in general and his practices 
in installing them in the houses that he 
builds. The wall-cabinet model was shown 
to him and his interest in the idea ascer- 
tained. Here again, an attempt was made 
at determining the degree of interest by 
requesting the respondent to make a guess 
as to how many he would have installed 
during the current year if they were avail- 
able. The final questions concerned the 
use of color and wood for kitchen cabi- 
nets — a growing trend at that time and 
one that would certainly have an effect 
upon the marketing of a wall-cabinet dish- 
washer. The demonstration of the wall- 
hung unit for the builders varied from the 
one for the consumers and the dealers in 
that the dimensions of the unit were spe- 
cifically discussed. 



As with the dealers, there was little 
problem concerning generation of interest. 
As a group, the builders considered the 
number and type of appliances installed 
in a home important to the sale of that 
home, since many felt that the kitchen 
was the most important room in the house 
for the housewife. Consequently, they 
were quite anxious to examine any new 
ideas in appliances and to voice an 
opinion. 

FINDINGS AND RECOMMENDATIONS 

The completed questionnaires were for- 
warded to New York every few days. Mar- 
shall's project examined them as they were 
received. Originally, it had been planned 
to extend the study west of the Missis- 
sippi. However, some fairly clear patterns 
of response developed and it was jointly 
decided by Continental and Marshall to 
limit the scope of the study to states east 
of the Mississippi. 

Upon the conclusion of the field work, 
a lengthy report was submitted to Conti- 
nental with findings and recommendations 
for each machine. Some of the highlights 
of the findings are as follows: 

Counter-Portable 

Only 17 per cent of the housewives in- 
terviewed preferred the counter-portable 
over the roll-around portable, and only 
seven per cent put it first among all four 
machines tested. There was no significant 
difference in preference among the groups 
who were expected to prefer this model: 
housewives with small families, renters, 
apartment house dwellers, and the lowest 
economic group interviewed. 

In addition, only 14 per cent of the 
dealers picked it over the roll-around, and 
only 22 per cent would consider stocking it 
(14 per cent of the Continental dealers 
interviewed). 



290 



EXPANDING DEMAND 



Criticisms of the machine voiced dur- 
ing the consumer interviews were "ca- 
pacity too small," "difficulty and nuisance 
of lifting/' and "loss of the valuable 
counter space that it would occupy." 

Roll-Around Portable 

Some 70 per cent of the housewives 
interviewed preferred the roll-around port- 
able over the counter-portable, and it 
was the first choice among the four ma- 
chines of 24 per cent of the housewives. 
There was higher-than-average preference 
among the young housewives, apartment 
dwellers, renters, and middle-class women. 

Dealers expressed moderate interest in 
handling this portable, whereas they pre- 
viously had showed little interest in port- 
ables in general. About 44 per cent of the 
dealers interviewed expressed a desire to 
carry a roll-around portable and the gen- 
eral consensus was that a roll-around in- 
cluded in the line for 1953 would have 
meant a 15 per cent net increase in busi- 
ness. 

Reasons listed for liking the roll-around 
were "does not require installation in space 
which would take away from a kitch- 
en cabinet," "maneuverable around the 
kitchen," and "can be taken along to a 
new home without trouble in the event 
the family moves." 



believe this would be additional business, 
but rather would be volume taken away 
from existing floor-counter models. 

Greater general interest in the wall-cabi- 
net model was shown by the builders than 
the dealers, but they, too, had qualifying 
reservations. The report indicated that 
about 10 per cent had unqualified interest 
in installing a wall-cabinet model; 38 per 
cent would install in higher-priced homes 
or in model homes or as an extra feature 
when housing gets hard to sell; 22 per cent 
were not interested but might be if de- 
mand' were built first (not willing to 
pioneer), and 30 per cent were not inter- 
ested at all. Interest appeared to be highest 
among eastern builders, medium-sized 
operators, those who already were install- 
ing dishwashers, and those now installing 
Continental appliances in their homes. 

Among the reasons stated for preferring 
the wall-cabinet model were "occupies" 
space more readily given up," "would be 
out of way where it wouldn't block floor 
or where small children wouldn't get at 
it," and "could be used as a storage space 
when not in use." Critical statements 
noted were "loading operations more diffi- 
cult when lifting upward than reaching 
downward," "appears to have less capacity 
than it's supposed to," and "it may leak." 
The lower price of the wall cabinet model 
was not raised as a factor in its favor. 



Wall-Cabinet Dishwasher 

Compared with the conventional floor- 
counter type of dishwasher, only 27 per 
cent of the housewives interviewed pre- 
ferred the wall-cabinet model. However, 
28 per cent of current dishwasher owners 
would have chosen this over what they did 
buy if it had been available. 

If stocked, the dealers felt that the wall- 
cabinet model would account for about 
20 per cent of their business. They did not 



Dishwasher Market in General 

Based on the consumer interviews, it 
was determined that the saturation level 
within the top 60 per cent of the economic 
scale was 6 per cent. 

Only 1 per cent of those interviewed 
had any definite plans about purchasing a 
dishwasher in the next six months, and 
less than 3 per cent had purchase plans 
for the next year. However, 22 per cent 
expressed interest in making a purchase in 



EXPANDING DEMAND 



291 



the indefinite future, and this figure rose 
to 30 per cent after exposure to the 
experimental units demonstrated in the 
trailer. 

Reasons expressed for the general lack 
of desire to buy were: "small family," 
"lack of room," "the expense," "want 
other appliances first," "dishwashers per- 
form poorly." 

Most dealers complained about compe- 
tition from discount and wholesale opera- 
tions and were not pushing dishwasher 
sales very aggressively. However, a majority 
of dealers predicated a moderate increase 
in sales in the near future. This was based 
upon an expectation that consumer accept- 
ance would be greater, dishwashers would 
become more standard equipment, there 
would be more installations in new homes, 
and manufacturers would stimulate the 
trend through intense promotion. 



Can Continental now make any deci- 
sions about new models that will help it 
to expand its sales of dishwashers, which 
it could not have made so clearly or con- 
fidently without the test described? What 
other kinds of research could it, or should 
it, do to find ways to improve product 
characteristics and thus increase the at- 
tractiveness of dishwashers to consumers? 
What compromise between the ideal and 
the practical in marketing research is ap- 
propriate here? 

APPENDIX 1 — QUESTIONS INCLUDED IN 
CONSUMER QUESTIONNAIRE 

Section A: Dishwasher Ownership and 
Awareness 

(Ascertain housewife's familiarity with 
dishwashers by asking the following ques- 
tions until a "yes" answer is obtained. 
Then move on to Section B or C.) 
la. Do you have an autmatic dishwasher 
in your home now? 



b. Have you ever used an automatic 
dishwasher? 

c. Have you ever seen an automatio 
dishwasher in operation? 

d. Have you ever seen an automatic 
dishwasher? 

e. Have you ever seen or heard anything 
about them either in newspapers, 
magazines, or TV or from friends or 
relatives? 

(If dishwasher is owned, ask Section 
B. If not, ask Section C.) 

Section B: Dishwasher in Home 

2. What make or brand is it? 

3. Is it portable or permanent? 

4. What type is it? (Show card illus- 
trating drop-door, pull-out and top- 
opening types) 

5. How many years old is it? 

6. Now that you've had your dish- 
washer a while, how do you like it? 
Why do you say this? 

Section C: No Dishwasher in Home 

7. Are you planning to buy a dish- 
washer in the future? 

a. If yes, how soon? 

b. If no, are there any particular rea- 
sons why you are not planning to 
buy one? 

Section D: Classification Data 

8. How many times a day do you wash 
your dishes? 

9. How many people live in your house- 
hold? 

10. What is your age? 

11. Record economic status. 

12. Determine whether home is owned 
or rented and type (one-family 
house, apartment, and so forth.) 

Section E: Mobile Unit Interview 
(Demonstrate both portables.) 

13. Which of these two dishwashers 
would you prefer for your kitchen? 
Why? (Probe carefully) 



292 



EXPANDING DEMAND 



14. Which of these two would you pre- 
fer for your kitchen? Why? 

(Now ask housewife to consider all 
four. ) 

15. Now that you have seen four types 
of dishwashers, including the port- 
able types and permanently installed 
types, which one of the four would 
you prefer for your kitchen? Why? 

16. If that machine (pointing to first 
choice) were not available, which of 
these three would you prefer for 
your kitchen? Why? 

17. If one of the portable models is a 
first or second choice, ask: Where 
would you keep the dishwasher when 
you weren't using it? 

18. If the floor-counter type is preferred 
over the wall type, ask: If you were 
buying a new home, which of these 
two would you prefer? 

19. Now that you've had an opportunity 
to look at various models and learn 
something about them such as how 
much they cost, how they load, and 
so on, do you think you'll be buying 
one in the future? 

a. If "yes," how soon? 

APPENDIX 2 — QUESTIONS INCLUDED IN 
DEALER QUESTIONNAIRE 

1. Type of outlet 

2. Position of person interviewed 

3. Location of dealer 

4. Types of dishwashers handled 

5. Brands carried 

6. How many dishwasher units (all types) 
did dealer sell in 1952? How many per- 
manents? How many portables? 

7. How many dishwasher units does he 
expect to sell in 1953? How many 
permanents? How many portables? 

8. Recap whether increase in 1953 over 
1952 or decrease or same. 

9. Why does he think 1953 will be as 
above? 



Section B: Future Dishwasher Potential 
1. What is your opinion as to the sales 
possibilities in the next few years of 
mechanical automatic dishwashers, gen- 
erally: 

Expect sales to increase considerably? 

Expect sales to increase moderately? 

Expect sales to remain the same? 

Expect sales to decrease? 
Why? 

Section C: Portables 

1. For your own operation, what is your 
opinion as to sales possibilities of port- 
able automatic dishwashers? 

Excellent? 

Good? 

Fair? 

Poor? 
Why? 
(Demonstrate portable.) 

2. If you were to sell a portable, which of 
these two would you prefer to sell? 
Why? 

3. How interested are you in handling 
portable dishwashers like these or even 
some other type of portable? 

Very interested? 

Mildly interested? 

Not interested? 
Why? 

(If dealer is not interested in portables, 
skip to Section D; otherwise continue.) 

4. Are you interested in carrying either 
or both of these? Why? 

5. Is there any other type of portable 
dishwasher other than these two that 
you would prefer to handle? Why? If 
"yes," describe. 

Ask questions 6, 7, and 8 for each type 
of portable (including any type dealer 
may suggest in Question 5) in which 
the dealer indicates an interest. Ask 
each separately, that is, first ask 6, 7, 
and 8 for one; and then 6, 7, and 8 on 
the other, and so forth. 
Record total number of units dealer ex- 



EXPANDING DEMAND 



293 



pects to sell in 19 S3 (from Question 7, 
Section A). 

6. Record number of units of each model 
dealer thinks he could sell in J 953, if 
he had been carrying it since January 
I, 1953. 

Counter portable? 
Roll-around portable? 
Other? 

7. Of these, record the number of units 
that would come from new customers 
who might not have purchased any 
dishwasher. 

8. Of these, record the number of units 
that would come from customers who 
might have purchased a permanent 
model. 

9. Why do you think people would want 
a portable? 

Section D: Wall-hung Dishwasher 

(Demonstrate wall-hung dishwasher.) 

1. For your own operation, what is your 
opinion as to the sales possibilities of 
a wall-hung permanent dishwasher? 

Excellent? 
Good? 
Fair? 
Poor? 
Why? 

2. How interested are you in handling a 
wall-hung dishwasher? 

Very interested? 
Mildly interested? 
Not interested? 
Why? 

3. Record number of wall-hung units 
dealer thinks he could sell in 19 S3, if 
he had been carrying it since January 
1, J953. 

4. Of these, record the number of units 
that would come from customers who 
might have purchased a floor-counter 
model. 

5. Record the number of units that would 
come from new custmers who might 
not have purchased any dishwasher. 



6. Why do you think people would want 
a wall-hung dishwasher? 

Section E: All Four Dishwashers 

Let us suppose that at the first of the 
year, you had been carrying both of 
these portables, a wall-hung dish- 
washer, and the present floor-counter 
type. How many units in each cate- 
gory do you think you would have sold 
in 1953? 

APPENDIX 3 — QUESTIONS INCLUDED IN 
BUILDER QUESTIONNAIRE 

Section A: Classification Data 

1. Name 

Person interviewed 

Position 

City 

State 

2. Type of builder (speculative or con- 
tract)? 

3. What price classes? 

4. What types of houses (ranch, Cape 
Cod, and so forth)? 

5. About how many houses did he build 
in 1952? 

6. How many does he expect to build in 
1953? 

7. Does he install dishwashers in homes 
he builds? 

Section B: Why Builder Doesn't Install 
Dishwashers 

1. Why is it that you don't install dish- 
washers? (Probe) 
Section C: How Builder Installs Dish- 
washers 

1. What brands do you install? Why 
these? 

2. In 1953, of the homes you will build, 
in about how many will you install 
dishwashers? 

3. If not 100%, why not in all homes you 
build? 

4. In what types or classes of homes does 
he install these dishwashers and why for 
each? 



294 



EXPANDING DEMAND 



Section D: Interest in Wall-hung Dish- 
washer 

(Give builder same type of demonstra- 
tion as given to consumers and dealers, 
plus dimensions of unit [30" x 13"].) 

1. What do you think of the idea of a 
wall-hung dishwasher? 

Excellent? 
Good? 
Fair? 
Poor? 
Why? 

2. If available, would you want to install 
them in homes you build? Why? 

(If builder is definitely not interested 
in a wall-hung dishwasher, terminate 
interview here. If he is, continue.) 

3. If available since January 1, 1953, how 
many would you have installed? 

4. How many would you have installed in- 
stead of floor-counter type? 



5. How many would you have installed in 
addition to the floor-counter type? 

6. Would you install a wall-hung in all 
types of homes, in all price classes, in 
all types of kitchens, or in particular 
types of homes? If particular types, 
what types and why? 

7. What material and color are kitchen 
cabinets that you usually install in 
homes you build? 

Material 

Steel 

Wood 

Other (Specify) 
Color 

White only 

Color (Specify) 

8. Would you want to have a wall-hung 
dishwasher match the kitchen cabinets? 
Why? (Ask builder for any comments 
he has.) 



EXPANDING DEMAND 



295 



CASE 4-3 

Hawaiian Pineapple Company 



THE PROBLEM 

There are three sources of evidence of 
fruit quality obtained by the Hawaiian 
Pineapple Company: (1) estimates of the 
value of fruit received by the factory from 
the plantation, sampled by fields; (2) lab- 
oratory measurements of quality, such as 
esters, acid, brix, and color; and (3) con- 
sumer distinctions and preferences. 28 The 
problem becomes one of relating consumer 
distinctions to known definitions by the 
laboratory and then, in turn, describing 
field production in these same laboratory 
terms. This would provide standards and 
a means of evaluating levels of factory ac- 
ceptance of the raw materials. 

Essential to such evaluation is the cali- 
bration of consumer distinction, since the 
laboratory can make refined distinctions 
which have no realistic bearing on con- 
sumer behavior. It also seems essential to 
us to supplement expert panels and opinion 
with a cross section of the public itself, 
mainly because differences, unimportant 
to the consumer, may otherwise be ac- 
centuated. 



THE EXPERIMENTAL PROCEDURE 

A project to test consumer reaction to 
fruit quality characteristics was initiated 
on May 28, 1952, and the general design 
of the experiment was formulated. The 



28 This case has been adapted from Raymond 
Franzen, "Development of Fruit Quality 
Standards on the Bases of Consumer Distinc- 
tions and Preferences," in Henry Brenner, 
Editor, Marketing Research Pays OS (Pleas- 
antville, N. Y., Printers' Ink Books, 1955), 
pp. 91-96. 



experimental procedure proposed was : ( 1 ) 
to measure the quality of samples of pine- 
apple, which is the raw material; (2) to 
produce from the laboratory samples which 
have given degrees of given chemical and 
physical characteristics found in the raw 
materials; and (3) to determine just 
which degrees of what are adequately 
distinguished by the consumer, and the 
direction of the consumer preferences, if 
any. 

The experiment was conducted in the 
San Francisco Bay area. Twelve large 
self-service stores handling Dole pineap- 
ple products were selected and interviews 
were conducted in February, 1953, on 
three successive week ends. 

In each store, a booth was erected ad- 
jacent to a special display of Dole prod- 
ucts and all incoming customers were 
presented with a coupon worth 5 cents 
on the purchase of a Dole product. If a 
customer selected a can of pineapple from 
the display, he was asked by the booth 
attendant to participate in the interviews. 
This selection of respondents was made 
to assure tests of people who do actually 
use pineapple. 

Each interview involved a taste in 
duplicate by the consumer. In this test, 
she first tasted sample X and then sample 
Y. In the duplicate test, the same con- 
sumer unknowingly re-evaluated the same 
samples, but sample Y was presented first 
and sample X second. There were nine 
such double comparisons evaluated by a 
total of 1,881 interviews, but each in- 
dividual made only one of these double 
comparisons. 



296 



EXPANDING DEMAND 



*fc & : 






8 «s^& 






o o 
o o 

O (M 



5» •§ ^ ~ g ?■> 



Q S 



£*.£ 



e co o) i j. 

** ■•» «e» ■*•» * 

S ^ [§ .co ^ 

e> © '§ co ** 



** 1.5 8 4. 



V ^ , ^ . CO 

•O ' N •S ' N , ^> !» ° 



<^ ^ •** -^ cu g 



JS O 









s *> S 

t S 8-S 



co .2 



OS 00 CO 

CNI CD Oi 

CM rH t-\ 



t- .2 
cm -e 



•* o 
co irs 

CM t-I 



> 


t-> 


.3 


a 


pa 


S 

2 


x 


"3 


bD 


Q> 


w 


S 



K £ 



PQ 



xi 
b£ 



.3 


T3 


*H 


O 


PQ 


U 


£ 


S 

A 


o 
*1 


o a 




o - s 


CO 


3 


> 


S£ 


.3 


8 § 


pa 


5 & 

co oi 


bp 





hD 


cu 


X 


s 


x: 
o 
PQ 


s 





o 






CD 






3 

'5 


8fi 






« a 


CO 


cu 


c s 


tr- 





co — 


<M 


X 
£ 


*" T3 




J2 QJ 




o 


03 (-( 

&2 




3 


C CO 




jD 


23 C 

to 




'C 


■J. o 




CO 


•° <D 




-5 


O X5 




TJ 


« ^ 




CD 


3 E 




"flj 


•- < 0) 


PQ 


o 

3 


^ o 
iJ E 


C 


o o 

•O in 


£ 


Q> cn 


o 




Ui ~^ 


J 


« .— 


-C 




& o 



s 



EXPANDING DEMAND 



297 





O 




fc 






£ 


lO 




as 


a 


* 


"5 

1 


o 
to 


to 
Cj 
£ 

to 


to 

^0 




■ex 


to 


«0 
3| 





8 $T 



• co 



-.sli 

j •» >k S (o o ** 

« 815 218 15 



• s © § to ** « 

S •" ^. S <o O) ■» 

^ u. •€!_ co !< S 

| S « o v S to 

s r to -ex 

s to 



w 

Q 

£ w 

CO 

O W 
W K 



Ig 

Q° 

co Z 
co o 

go 

^* 

w 



g 



to Sf' to O -5 

co -5 58 to T3 

O co to to ^ 

^ 10 K 10 g 

to «. S ^ 



& *f !5 g 



§ <X SS 



k« to -^ £ 

to SP I ? * 

" °j ** ^ ~ 

p\\ t\\ f\\ S: 



J* £^.£ to 






s, « 



Kto^>to£>to£:s 

* s •*■ § $ s «. 

•^Js^Sco^toO 






CO 

•ex 



Q 8 S 



.3 
'u 

CO 


.3 

u 

pq 


.3 
PQ 


xi 
bp 


T3 
CD 

S 


bp 


© 
in 


CO 

in 


O 


PQ 


.a 

pq 


.3 


xi 
bp 


<d 


xi 
bD 


■<r 




CO 

in 



.a 

PQ 


.a 

PQ 


.3 

PQ 




T3 
CD 

s 


J3 
bp 

X 


IfS 


CO 
CO 


co 

CO 


•3 
PQ 


.3 

PQ 


.a 

u 

PQ 


Xi 
bp 


T3 


us 
bD 


in 

in 


C30 

in 


CO 
CO 



co 


i-t 


CO 


T-* 


t- 


Tf 


m 


o> 


CO 


o 


iH 


C- 


CO 


Tf 


•H 


H 


CM 


CM 


r-t 


.-» 


l-H 



m co 
o co 

CM ^H 



"* O 00 

co in co 

CN i-4 CM 



CJ1 C- 

m m 
i-H eM 



t- 


CO 


CO 


o> 


CO 


o 


o 


CO 


c- 


c- 


(M 


CM 


i-H 


t-H 


CSJ 





.3 

Sh 

PQ 


.3 
PQ 


.3 




.3 
PQ 

a 


.3 
u 
PQ 


.3 

u 


73 








a 

CO 




3 


o 


PQ 




2 


o 


PQ 


U 








"" 


1 




CO 

> 


o 


-<-• 


cu 

S 


co 
> 


1* 

3 


o i; 


.3 


.a 

u 
PQ 


.a 


22 

o 
E 


a 

o 


CO 

> 


'u 


CO 

> 


"3 


CO* 

> 




co 

> 


" 3 


PQ 


a 

3 


PQ 


o 
Ifl 


3 


.3 


PQ 


.3 


C1J 


.3 


PQ 


.3 


Is 


Xi 
bo 


'•& 


i* 


cj 


CO 


PQ 


a 


PQ 


! 

o 

5 


PQ 


a 

3 


u 

PQ 


S o 
h2 


K 


CD 

s 


o 


L 




bo 


1 


bD 


bp 


cu 


xi 
bfi 


1 


1 


1 


* 



298 



EXPANDING DEMAND 



It is possible to vary the order of the 
original pair: to have a progression of 
YX to XY, as well as XY to YX, but 
in this study, this was not done. A pref- 
erence was solicited in accordance with 
the sample questionnaire shown below. 
It should be pointed out that this method 
does not count a preference unless a valid 
distinction has first been made. It is pos- 
sible that distinctions made in the ex- 
periment are somewhat more frequent 
than actual distinctions made in the home, 
but this increases the rigor of the criterion. 



No Preference 



(Please 



check if no preference indicated.) 
3. Did customer purchase pineapple? 

Yes No 

The laboratory criteria were so manipu- 
lated that we had nine paired samples 
which may be seen in Table 4-L and 4-M. 
They use three possible comparisons of 
brix level for opaque fruit and three for 
translucent fruit and a pairing of opaque 
with translucent fruit at each level of 
brix. 29 



QUESTIONNAIRE 

Interviewer's Name 

Store Name and Location 
Date 



1. Before you are two kinds of pineapple. 
Will you please taste them and tell me 
which one you like better? 

Prefer (Please write in 



product number.) 
No Preference 



— (Please 

check if no preference indicated.) 

2. Now here are two others. Will you 

please taste them and tell me which 

one you like better? 

Prefer (Please write in 

product number. ) 



What questions can be answered by 
this test? 

What findings emerge that will be use- 
ful to the management of the company? 
In making what decisions? In what other 
kind of situations can the approach de- 
scribed here be used? 



29 The terms "high," "medium," and "low" 
brix refer to fresh fruit. All samples were 
canned in sugar syrup which resulted in ap- 
proximately the same sweetness. The posi- 
tions "high" and "low" brix are approxi- 
mately 1 from the mean of an average supply 
of fruit. Translucent fruit is fruit with low 
acid and high esters, while opaque fruit is 
fruit with high acid and low esters. 



EXPANDING DEMAND 



299 



CASE 4-4 

Holiday House 



Holiday House is one of the leading 
specialty mail-order firms in the U.S.A. 
The primary source of sales for the gift 
line is the 64-page Christmas gift catalog. 
Over a million copies are mailed annually. 
Management's constant concern is to in- 
crease the sales effectiveness of the Chris- 
mas catalog. Management normally could 
forecast Christmas catalog sales within 
± 5% of final results. Over time, how- 
ever, management's estimates became con- 
siderably less reliable. This factor, coupled 
with increasing competition in the mail- 
order field, indicated the necessity to find 
more accurate methods of forecasting sales 
as well as methods to increase sales and 
reverse a downward trend. 

Management, therefore, decided to con- 
duct a thorough scientific marketing re- 
search study of its entire operation in the 
expectation that the ability to forecast 
sales would be sharpened, and a better 
understanding of the nature of the mail- 
order business would turn up new ways to 
increase sales. 

The marketing research group engaged 
to conduct this study decided to use 
personal interviews, telephone interviews, 
and large-scale mail questionnaire sam- 
plings among mail-order buyers on a na- 
tion-wide basis. 

Pre-test results quickly indicated that 
while a great deal of new information 
would be gathered by these methods, not 
much information regarding the sales ef- 
fectiveness of the catalog itself would be 
revealed. Strong but inconclusive feelings 
about the catalog were detected. Indica- 
tions were that even minor improvements 
in the catalog and method of merchandise 
presentation could be very effective in 



the converting of nonbuyers into buyers. 
It was also apparent, however, that 
interviewees could not specify exact feel- 
ings or reactions to the catalog, because 
the aesthetic considerations of art, color, 
typography, printing, etc., became too 
complicated. Also, it was found that most 
catalog buyers were quite fond of catalog 
shopping and tended to report only favor- 
able feelings about all catalogs they had 
seen. The researchers decided, therefore, 
to conduct a searching scientific analysis 
of company internal sales data based on 
the sales results achieved by each page of 
the last four catalogs in order to deter- 
mine the extent to which various factors 
influenced the sales results of each page 
in these catalogs. Excerpts from the re- 
searchers' report are presented in the 
Appendix. 



What conclusions and/or recommenda- 
tions are justified on the basis of an 
analysis of the following appendix and 
exhibits? Which variables are important? 
How much improvement in sales can be 
expected if your recommendations are fol- 
lowed? What other kinds of marketing 
problems can be solved by this approach? 



APPENDIX — ANALYSIS OF HOLIDAY HOUSE 
SALES DATA ON PAGES IN CATALOG 

The first analysis of company sales data 
was focused on an examination of the 
factors which might have determined net 
sales per page. It was decided to begin 
with sales per page rather than item profit 
per square inch, because the former 
involved 216 pages whereas the later in- 



300 EXPANDING DEMAND 

TABLE 4-N. VARIABLES ON WHICH SALES PER PAGE WERE CODED AND ANALYZED 



Column (on punched cards) 

1,2,3 Page rank in order of 

gross margin per square 
inch. 

4 Year 1955 - 5* 

1956 - 6 

1957 - 7 

1958 - 8 

5,6 Number of the page in 

the catalog. 

7,8,9 Gross margin per square 
inch in hundreds of 
dollars. 

10,11,12 Dollar sales per page 
in hundreds of dollars. 



Column 



20 



21 



22 



13,14 


Number of items per page 


15 


Number of items under $2 


16 


Number of items $2- $5. 


17 


Number of items $5-$10. 


18 


Number of items over $10 


19 


Number of imports per 




page. 



23,24,25 



26 



27 



28 



Color 




None 





Green tone 


1 


Red 


2 


Mint 


3 


Orange 


4 


Pink 


5 


Yellow 


6 


Use of color 




None 





Borders & captions 


1 


Background 


2 


Functional 


3 


Tone 


4 



Orderliness of Layout 

Orderly 1 

Fair 2 

Jumbled 3 

Messy 4 

Page rank in order of dollar 
sales. 

Quartile page ranking in 
order of dollar sales in each 
catalog year. (1955,1956,1957, 
1958, each year considered 
separately) 



1st quartile 
2nd quartile 
3rd quartile 
4th quartile 



(highest) 



(lowest) 



Number of types of item on page. 

One 1 

More than one 2 

Number of items per page. 

2-3 

4-5 1 

6-7 2 

8-9 3 

10-11 4 

12-13 5 

14-15 6 

16-17 7 

18-19 8 

20-21 9 



♦Each page was coded on an IBM card. These cards each contain eighty columns of which only 
columns 1-29 were used. In each column, it is possible to use punches to indicate up to twelve 
different pieces of information; a blank, or no punch, adds a thirteenth. For example, in column 
4, a 5 punch would indicate 1955. Punching in column 5, and 1 in column 6 would indicate page 
01 (i.e., page 1) of the catalog. In column 29 there were thirteen categories of answers. 



EXPANDING DEMAND 



301 



Column 
29 



TABLE 4-N (Continued) 



Type of items 

Blank Non-classifiable 

Personal accessory, houseware, and novelties 

1 Christmas decorations only 

2 Toys only 

3 Auto accessory and hardware 

4 Housewares and kitchenwares 

5 Housewares, kitchenwares, and novelties 

6 Personal accessories and housewares 

7 Novelties only 

8 Personal accessories and jewelry 

9 Toys and novelties 

10 Personal accessories and toys 

11 Personal accessories and novelties 



volved about 2,500 items for the Christ- 
mas catalogs, 1955-1958. It was realized 
that a detailed analysis of items profit per 
square inch might be necessary in order 
to obtain more complete information, but 
the page analysis was done first to test this 
approach. 

There were 216 pages in the Christmas 
catalogs of the years 1955-1958, exclud- 
ing cover pages. Each page was examined 
and certain variables were coded onto 
IBM cards. Table 4-N is a list of the 
variables that were coded for each page 
in the catalog. All the variables pertain- 
ing to sales or gross margin per page were 
obtained from the company records and 
all other variables pertaining to charac- 
teristics descriptive of a page were as- 
signed by the research group. 

When characterizing a page as to the 
type of merchandise on that page, a cer- 
tain amount of judgment was involved, 
because only a few pages were made up of 
only one category of merchandise. It was 
necessary, therefore, to define merchandise 
categories that covered the common com- 
binations of merchandise (Column #29, 
Table 4-N). This combination resulted 
in uncertainty as to exactly how much of 
any one page consisted of a particular 
type of merchandise. An item-by-item 
analysis rather than a page-by-page analy- 
sis would have been necessary to eliminate 
this difficulty. 



In attempting to see whether particular 
variables had any effect on sales per page, 
the pages were first ranked in deciles ac- 
cording to sales volume. The decile rank- 
ing showed some differences between 
pages with high sales and those with low 
sales but was hard to work as the result- 
ing tables were cumbersome and cell sizes 
became very small. Therefore, the pages 
were then ranked by quartiles on sales 
volume per page within each year. Yearly 
sales quartiles were used to offset the ef- 
fect of different catalog performances in 
different years. For purposes of simplifica- 
tion, it was then decided to further group 
the quartiles into first and second halves 
in order to facilitate visual examination 
of the results. 

In Tables 4-0 through 4-U the effects 
of seven different factors are presented. 
These factors are position in catalog, 
number of items per page, merchandise 
type, number of imported items, color, 
use of color, and layout orderliness. In 
each of these exhibits separate data are 
presented for each year, 1955-1958, and 
for the four years combined. For example, 
referring to Table 4-0, we see that for 
the four years combined there were 30 
pages which contained 8 or 9 items and 
that 11 of these pages had above median 
sales, while 19 had below median sales. 
Each exhibit also shows the trend, over 
time, in the factor being analyzed. 



TABLE 4-0. EFFECT OF POSITION IN CATALOG ON SALES 
PER PAGE 

Sales Results Shown by Yearly Ranking-lst or 2nd Half 

Page in Catalog 





1-9 


10-19 


20-29 


30-39 


40-49 


50-59 


Total 


1955-58 
















1st half 


21 


24 


22 


19 


15 


7 


108 


2nd half 


15 


16 


18 


21 


25 


13 


108 


1955 
















1st half 


5 


5 


5 


5 


5 


3 


28 


2nd half 


4 


5 


5 


5 


5 


4 


28 


1956 
















1st half 


4 


6 


9 


4 


3 


2 


28 


2nd half 


5 


4 


1 


6 


7 


4 


27 


1957 
















1st half 


7 


6 


5 


5 


3 


2 


28 


2nd half 


2 


4 


5 


5 


7 


5 


28 


1958 
















1st half 


5 


7 


3 


5 


4 





24 


2nd half 


4 


3 


7 


5 


6 





25 






Trend in 


Page Position in 


Catalog 






1955 


9 


10 


10 


10 


10 


7 


56 


1956 


9 


10 


10 


10 


10 


6 


55 


1957 


9 


10 


10 


10 


10 


7 


56 


1958 


9 


10 


10 


10 


10 





49 


Sort Code Columns 


: * 
















26-5 
















4-26- 


■5 














4-5 













♦Instructions for machine sorting. 



TABLE 4-P. EFFECT OF NUMBER OF ITEMS ON PAGE ON SALES PER PAGE 

Sales Results Shown by Yearly Ranking-lst or 2nd Half 

Number of Items Per Page 



Total 


2 &3 


4 &5 


6 &7 


8 &9 


10 & 11 


12 & 13 


14&15 


16 & 17 


18 & 19 


20 &21 


Total 


1955-1958 
























1st half 





1 


1 


11 


36 


29 


15 


10 


2 


3 


108 


2nd half 


1 


1 


3 


19 


24 


32 


15 


9 


2 


2 


108 


1955 
























1st half 











6 


15 


4 


2 


1 








28 


2nd half 











4 


10 


9 


2 


3 








28 


1956 
























1st half 














6 


13 


6 


2 


1 





28 


2nd half 





1 


1 


2 


5 


12 


5 


1 








27 


1957 
























1st half 





1 


1 


3 


13 


4 


2 


2 





2 


28 


2nd half 








2 


11 


5 


4 


3 


3 








28 


1958 
























1st half 











2 


2 


8 


5 


5 


1 


1 


24 


2nd half 


1 








2 


4 


8 


5 


2 


1 


2 


25 








Trend in Number of Items Per Page 4 










1955 











10 


25 


13 


4 


4 








56 


1956 





1 


1 


2 


11 


25 


11 


5 


1 





55 


1957 





1 


3 


14 


18 


8 


5 


5 





2 


56 


1958 


1 








4 


6 


16 


10 


7 


2 


3 


49 


Sort Code : 


26-13 


4-26- 


13 4- 


-13 

















EXPANDING DEMAND 



303 



w 5 

co eg 



3 



CO . — 



<0 



t to 






£1 



cj 3 






c"3 H 



is J< 

£2 



co ••*> 












si 



CO ~i 

CO .© 



csj co 
CO co 

00 10 



CO CD 



c^c^ t^c^ S^S^ 

OO OO OO 

o 00 cm © 



cm © ^ i-t 



m m 



HN t-I 



o 



o o 

m in 



© o 
o 



es 



o © 



egg 



CM CM 



O O 
CM 00 



O 1-1 



cm in *-« in 



CM O O i-i 



O 
O »H OO 

CM t- 

<N ^ r- •<* cm in 



to tt co t- Tf co 

coco C0«O coco Nt- 



00 00 

tT CO O 



t- CO 00 

CD CO 00 CM 



C^i^ t^t^ c^^ 

00 00 00 
000 



o o 

CO ^» CO CO 



f?S? 



! f 



es^s e^^ e^t^ c^e^ ^^ 

00 cm eoc-v 00 cot- co r- 

coco coco o coco toco 



B£c* 
O O 
to ^ 



CO t- 

cm tr- 



00 CM 

CO to 



to ^ 

CO CO 



t- CO 

in ■«* 



00 CO 

co to 



t- CO 
C- CM 



in in 
c- CM 



C- CO 

in ^« 



gg 



t^c^ 

t- CO 
t- CM 



c» 

•f^. in in co cm 



1 

§ 

•« eo Tf to 

"2 Tf CO Tj< 



gg gc^ 



co m ^h 



t*?BR 
co t- 
CO to 



c^l^ 

t- CO 



<3> 

1-1 00 



e*5R 
in in 
cm t» 



rj« to 
CM CO 



#?^ 



gg gg gg gg gg 



ao 



«* 



o 

O i-t 



o 

i-H 
CO O 



e* 



o 

CO O 

© o 



© 

1—1 

«-H O 



o o 
© 



C- "«* CM CM 



00 o> o» in 



co co ^ co 



i-H co CO i-4 



CO CM i-4 tM 



* 5 



o <r> 00 c (cue ^ouc 

C*« "1 1^1 PN 







•O r* CM 


00 CO c 
in ^ cm 



S 



u 

WCOf CO v 

in m in m u 
o> 01 o> o> o 

,-t H T-I 1—1 CO 



TABLE 4-R. EFFECT OF NUMBER OF IMPORTS PER PAGE ON SALES PER PAGE 
Sales Results Shown by Yearly Ranking— 1st or 2nd Half 
Number of Imports Per Page 

























10 or 









1 


2 


3 


4 


5 


6 


7 


8 


9 


more 


Total 


Total 


























1955-1958 


























1st half 


42 


21 


20 


11 


5 


6 


1 


1 








1 


108 


2nd half 


36 


18 


24 


8 


10 


7 





3 


2 








108 


1955 


























1st half 


11 


3 


8 


2 


3 


1 

















28 


2nd half 


9 


6 


12 





1 




















28 


1956 


























1st half 


11 


8 


6 


1 





1 


1 














28 


2nd half 


9 


4 


5 


3 


2 


2 





2 











27 


1957 


























1st half 


13 


5 


3 


3 


1 


2 





1 











28 


2nd half 


8 


6 


3 


2 


4 


4 





1 











28 


1958 


























1st half 


7 


5 


3 


5 


1 


2 














1 


24 


2nd half 


10 


2 


4 


3 


3 


1 





2 











25 








Trend in 


Number Imports Per 


Page 










1955 


20 


9 


20 


2 


4 


1 

















56 


1956 


20 


12 


11 


4 


2 


3 


1 


2 











55 


1957 


21 


11 


6 


5 


5 


6 





2 











56 


1958 


17 


7 


7 


8 


4 


3 








2 





1 


49 



Sort Code: 26-19 



4-26-19 



4-19 



TABLE 4-S. EFFECT OF COLOR ON SALES PER PAGE 
Sales Results Shown by Yearly Ranking- 1st or 2nd Half 





Single Color 






Two ' 


Color 












Green 




Black 


Black 


Black 


Black 


Black 




Combit 




Black 


Tone 


Total 


&Red 


& Mint 


& Orange 


&Pink 


& Yellow 


Total 


Total 


Total 






















1955-1958 






















1st half 


42 


11 


53 


33 


9 


4 


5 


3 


54 


107 


2nd half 


39 


5 


44 


43 


8 


4 


5 


5 


65 


109 


1955 






















1st half 


12 


5 


17 


11 














11 


28 


2nd half 


12 


3 


15 


13 














13 


28 


1956 






















1st half 


13 


6 


19 


9 














9 


28 


2nd half 


11 


2 


13 


14 














14 


27 


1957 






















1st half 


17 





17 


11 














11 


28 


2nd half 


16 





16 


12 














12 


28 


1958 






















1st half 











2 


9 


4 


5 


3 


23 


23 


2nd half 











4 


8 
Trend in 


4 
Color 


5 


5 


26 


26 


1955 


24 


8 


32 


24 














24 


56 


1956 


24 


8 


32 


23 














23 


55 


1957 


33 





33 


23 














23 


56 


1958 











6 


17 


8 


10 


8 


49 


_49 
216 


Sort Code: 


26-20 


4- 


26-20 


4-: 


20 













304 



TABLE 4-T. EFFECT OF USE OF COLOR ON SALES PER PAGE 



Total 

1955-1958 
1st half 
2nd half 

1955 
1st half 
2nd half 

1956 
1st half 
2nd half 

1957 
1st half 
2nd half 

1958 
1st half 
2nd half 



Sales Results Shown by Yearly Ranking- 1st or 2nd Half 
Borders 
Black & 

Only Captions Background Functional Green 



42 
39 

12 
12 

13 
11 

17 
16 






39 
41 

11 
13 

9 
14 

11 
12 



16 
23 










16 
23 



Total 



108 
108 



28 
28 

28 
27 

28 
28 

24 
25 



1955 
1956 
1957 
1958 



24 

24 

33 





Trend in Color Usage 

24 

23 

23 

10 39 



56 
55 
56 
49 



Sort Code: 26-21 



4-26-21 



4-21 



TABLE 4-U. EFFECT OF ORDERLINESS OF LAYOUT ON SALES 

PER PAGE 

Sales Results Shown by Yearly Ranking-- 1st or 2nd Half 



Total 

1955-1958 
1st half 
2nd half 

1955 
1st half 
2nd half 

1956 
1st half 
2nd half 

1957 
1st half 
2nd half 

1958 
1st half 
2nd half 



Orderly 



49 
42 

24 
19 

19 
19 

6 
4 






Fair 



39 

37 

4 
9 

9 
6 

21 
19 

5 
3 



Jumbled 



13 
20 






1 



1 
5 

12 
14 



Messy 



Total 



108 
108 

28 
28 

28 
27 

28 
28 

24 
25 



1955 
1956 
1957 
1958 



Trend in Orderliness of Layout 

43 13 

38 15 1 1 

10 40 6 

8 26 15 



56 
55 
56 
49 



Sort Code: 4-22 



4-26-22 



26-22 



305 



306 



EXPANDING DEMAND 



The International Latex Corporation 
manufactured various dipped latex and 
related products. The company's 200 sales- 
men were divided into two main divisions, 
and sold its products direct to thousands 
of retail outlets, under the brand name 
"Playtex." International Latex was an 
important manufacturer of waterproof in- 
fants' pants, rubber kitchen gloves, bath- 
ing caps, foam pillows, girdles and bras- 
sieres. The company's record had been 
one of steady growth in sales, earnings, 
and breadth of product line. Sales had 
increased from less than $4 million in 
1946 to more than $32 million in 1955. 

Latex habitually spent a large percent- 
age of its sales upon consumer advertising 
in the leading magazines and Sunday 
supplements, as well as maintaining a 
substantial cooperative advertising progam. 
In 1955, the advertising budget was $7 
million as compared to sales of about 
$32 million. 

By 1956, television had established it- 
self as one of the most effective mass 
advertising media. The 600,000 TV sets 
in 1946 had grown to 34 million sets by 
the end of 1955. In early 1956, the execu- 
tives of Latex were considering negotiating 
a contract for ten one-minute TV spot 
advertisements per day every day for five 
years in a minimum of the top 100 mar- 
kets in the United States. Such a cam- 
paign would be the largest spot television 
campaign in TV history, and the TV time 
would be valued at $22 million annually. 
Because of the size of the financial com- 
mitment, and the uniqueness of TV ad- 
vertising for its particular product line, 
Latex directed its research department to 



CASE 4-5 

International Latex Corporation 

test the effect of TV advertising on Play- 
tex brassiere and girdle sales. 

Playtex girdles had been an established 
product for many years, while Playtex 
brassieres were just out of the introduc- 
tory stages. 

SALES RESULTS TEST 

The Latex marketing research depart- 
ment in conjunction with its advertising 
agency, Foote, Cone and Belding, decided 
to test the campaign by measuring sales of 
stores in a TV test market against sales 
of stores in control cities which had no 
Playtex TV advertising. To have attempted 
to isolate TV as an advertising medium 
from TV copy and point-of-sale mer- 
chandising would have been not only 
time-consuming but also very expensive. 
Furthermore, it would be television ad- 
vertising which would make the increased 
merchandising effort possible. For example, 
a department store would he more apt 
to promote actively a girdle that was 
heavily advertised on TV than one which 
was not. Consequently, the objective of 
the sales results test became: to determine 
the effect of TV advertising and its re- 
lated merchandising program on the sales 
of Playtex girdles and brassieres in a 
specific test market. 

The company decided to use metropoli- 
tan New York as the test area. Latex had 
already decided to make New York its 
first market if it went ahead on TV, so 
the test was looked upon somewhat as a 
head start for its first campaign. The com- 
pany's national sales office was in New 
York City, and the matter of convenience 



EXPANDING DEMAND 



307 



entered into the decision. New York, one 
of two seven-channel cities in the United 
States, was generally conceded to be the 
toughest TV market in the country. Suc- 
cess in New York would be of much more 
value from a promotional viewpoint than 
success in a lesser, easier market. It should 
be emphasized that New York was not 
selected because it was a representative 
market, nor was there to be any attempt 
at projecting the test sales results on a 
national basis. 

Hartford, Rochester, and Philadelphia 
were selected as control cities. Geographi- 
cally, they surrounded New York, yet they 
were close enough to minimize regional 
variations. Moreover, each of the cities 
was large enough to have department 
stores comparable to those in New York. 
A significant portion of Playtex girdles 
and brassieres was sold to large depart- 



ment stores so that it was essential to 
have them represented in the survey. 

In each of the four areas, representative 
accounts were selected to be included in 
the tests. A leading department store 
as well as several specialty shops were 
picked in each control area. The ac- 
counts selected had to be cooperative 
ones, so there was no attempt to make a 
random selection. As seen in Table 4-V, 
there were 21 accounts in the control 
cities, and 31 accounts scattered through- 
out four boroughs, Westchester, and Ne- 
wark in the New York test market. 

Three one-minute TV spots per day 
were run in New York, Monday through 
Friday for four weeks starting Monday, 
April 2. Two of the spots were on WABC- 
TV. The first of these was scheduled be- 
tween 3 p.m. and 5 p.m. on the "After- 
noon Film Festival," and the second be- 



TABLE 4-V 
INTERNATIONAL LATEX CORPORATION 



New York 

Bloomingdale's Department Store, Manhattan 

Abraham and Straus, Brooklyn 

B. Gertz Inc., Jamaica 

Kresge, Newark 

small accounts in Bronx, Manhattan, Brooklyn, 

Queens, Westchester and Newark 
Total of test market accounts 

Rochester 

Sibley Lindsey and Curr 
small accounts 

Hartford 

G. Fox 

small accounts 

Philadelphia 

N. Snellenburg and Co. 

small accounts 

Total of control area accounts 



Number of 
accounts 

2* 

2 

2 

1 

24 
31 



2 
_5 
21 



*Latex classifies separate departments of large stores, such as notions 
and corsets, as separate accounts. 



308 



EXPANDING DEMAND 



INTERNATIONAL LATEX CORPORATION 

TELEVISION COMMERCIAL 



Product 


: Bra 




Length: 


one m 


inute 


Ident.: 


Tennis 




Code No 


. : B-2-56-1 






Friend: Wow! Set po 



Marvelous form, Carol. What's you 
secret? 



Girl: One you'd never suspect. Look. 




For top form I play in Playtex 




Playtex Living Bra is heavenly 
comfort for action all day long. 




This exclusive all-elastic design never 
slides, rides or shifts. 




My Playtex low-cut back stays low 
— never rides up and these exclusive 
elastic bias side panels adjust to 
every movement. 




Elastic criss-cross front dips low 




All elastic top never cuts or binds 
No broken straps . . . 




Glorious too for casual clothes 



'UIJIIJHH 




Or loveliest dr 




Try on a Playtex Living Bra at your 
store. You'll love it. Only $3.95. 
Wear Playtex Living Bras once — 
you'll never wear any others. 



Figure 4-W. International Latex Corporation — Television Commercial 



EXPANDING DEMAND 



309 



tween 7 p.m. and 10 p.m. The third spot 
was on WCBS-TV during the "Late 
Show" which ran from 11:15 p.m. until 
about 1 a.m. Equal time was devoted to 
the advertising of girdles and brassieres. 
However, in an attempt to obtain greater 
initial impact, all the spots during the first 
week were on girdles, while those of the 
second week were on brassieres. During the 
third and fourth weeks, the products were 
alternated on a spot-by-spot basis. A rep- 
resentative spot is shown in Figure 4-W. 

It was realized that factory shipments 
were a poor and often misleading indica- 
tion of store sales. Consequently, the re- 
search department decided to base its store 
sales figures on inventory figures adjusted 
for purchases. 

These sales figures were to be obtained 
by the regular Playtex salesmen who nor- 
mally called on the accounts. In the de- 
partment and specialty store field there 
were no services comparable to Nielsen 
(see Case 5-4). Moreover, department 
stores were not generally cooperative in 
giving their sales figures to outside agen- 



cies. 



Following a long-established company 
policy, Playtex salesmen had created a 
working relationship with almost all their 
accounts which permitted them to count 
the inventory on their regular sales calls. 
Thus, the salesmen not only were expen- 
enced at taking inventory, but also were 
familiar with the location of all the Play- 
tex stock. To stimulate accurate reporting, 
it was decided to give each salesman an 
extra $25 as compensation for his effort. 
With this as a background, Latex thought 
its salesmen were capable of doing a more 
accurate inventory check than employees 
of an outside agency. 

It was felt that the salesmen would do 
accurate work, and would not bias the re- 
sults, if they completely understood the 
survey. The district sales managers of each 



of the four areas, as well as the eastern 
regional sales manager, met in New York 
with the research director and the vice 
president of sales. The objectives, impor- 
tance, and details of the tests were 
thoroughly explained. The meeting also 
covered the uses of the point-of-sale ma- 
terial and the salesmen's extra compensa- 
tion. The district managers were made 
responsible for instructing the salesmen 
involved and supervising their activities. 

Salesmen took inventory in their test 
accounts two weeks prior to the start of 
TV. They again checked inventory just 
before the TV, adding in any merchandise 
received during the period, and then 
calculated the sales for the two-week 
period. The resulting figure was broken 
down into the average sales in units per 
day for each test account. These figures 
for each test account in the three con- 
trol cities were added together, and the 
total was given the base index of 100. 
Similar calculations were made for the 
31 accounts in the New York market, and 
a sales index of 100 was established. The 
salesmen checked the inventory and cal- 
culated sales in the stores two weeks 
after the start of TV, two weeks later as 
the TV ended, and two weeks after the 
end of TV. (See Figure 4-X.) Thus, they 
arrived at sales figures for three periods 
which were translated into index figures. 
The results of the sales test are also in- 
cluded in Figure 4-Y. 

Interpretation of the girdle sales figures 
is more complicated. The last two control 
area sales figures are somewhat biased by 
the fact that a new Playtex girdle at 
$5.95 was introduced in the control city 
of Philadelphia by error. The new girdle 
was also introduced in New York near 
the end of the final selling period. Be- 
cause of Latex's record of highly suc- 
cessful introduction of new products, their 
accounts actively promoted the new prod- 



310 



EXPANDING DEMAND 



ucts in an attempt to get an increasing 
share of the plus business. Consequently, 
Latex realized that the third and fourth 
index figures for the control areas were 
slightly inflated, and that the fourth fig- 
ure for New York was inflated somewhat 
more. 

HOOPER SURVEY 

Latex decided to do some further sur- 
veying as a check on the indications of 
the store sales test. Specifically, the com- 
pany wanted to know the influence of its 
TV advertising on brand awareness, pur- 
chase consideration, and actual purchases 
of Playtex girdles and brassieres. 

C. E. Hooper Inc. was engaged to make 
a series of telephone surveys of women. 
It was decided to do three surveys. The 
first was a random sample of 1,000 of 
the listed telephones in the New York 
metropolitan area; this survey was made 
two days before the start of the TV ad- 
vertising. Second, a survey identical to the 
first one was made with a fresh sample 
of 1,000 at the end of the TV advertis- 
ing. Third, from the first list of 1,000, 500 
were picked at random and also surveyed 
two days after TV ended. 

The size of the lists was determined 
by Hooper's past experience in the 
number of "don't answers," and the fact 
that the list had to produce 1,000 com- 
pleted interviews with women. The sur- 
vey was conducted by Hooper's experi- 
enced female telephone interviewers. Most 
of the interviews were completed during 
the day, but in an effort to eliminate any 
substantial bias against the working women, 
some interviews were made at night. If 
the prospective respondent was out on 
the first call, the Hooper interviewer went 
on to the next name on the list. 

The two questionnaires used are shown 
in the appendixes. Questions 1, 2 and 3 



dealt respectively with brand awareness, 
purchase consideration, and actual pur- 
chases of girdles. Questions 4, 5, and 6 
dealt, similarly, with brassieres. The first 
questionnaire was used on both of the 
groups of 1,000 respondents. 

APPENDIX — FIRST AND THIRD HOOPER 
SURVEYS 

This is the General Research Bureau 

calling. May I speak to ? 

We're making a survey on women's cloth- 
ing. Will you answer some questions for 
me? 

1. (a) Can you give me the names of 
several brands of girdles? 

(b) Can you name any others? 

2. The next time you plan to buy a 
girdle, what brands would you consider 
buying? 

3. (a) Have you bought any girdles in 
the past four weeks? 

If "no," skip to 4. If "yes": 

(b) What brand or brands did you 

buy? 

4. (a) can you give me the names of sev- 
eral brands of brassieres? 

(b) Can you name any others? 

5. The next time you plan to buy a 
brassiere, what brands would you con- 
sider buying? 

6. (a) Have you bought any brassieres in 
the past four weeks? 

If "no" end the interview. If yes: 
(b) What brand or brands did you 
buy? 
Thank you very much. 

APPENDIX — SECOND HOOPER SURVEY 

May I speak to ? This is 

the General Research Bureau calling 
again. You very kindly answered some 
questions on women's clothing for me 
about a month ago. May I ask you those 
questions again? 



Figure 4-X 

International Latex Corporation 

SALES CALCULATED AT TWO-WEEK INTERVALS 



March 19 



Four Weeks TV Advertising 
April 2 April 16 April 26 



May 14 



Base Sales = 100 

A 



First Index 

B 



Second Index 

c 



Third Index 

D 



TABLE 4-Y 
INTERNATIONAL LATEX CORPORATION 

Sales Text Indices 



Girdles 


A 
Two weeks 
before TV 


B 

First 

two weeks 


C 

Second 

two weeks 


D 
Two weeks 
after TV 


Philadelphia 

Rochester 

Hartford 


100 


101 


121* 


117* 


New York (TV 


) 100 


139 


127 


176* 


Brassieres 










Philadelphia 

Rochester 

Hartford 


100 


87 


88 


62 


New York (TV 


) 100 


130 


116 


128 



♦New $5.9 5 Playtex girdle introduced in error. 

TABLE 4-Z 
INTERNATIONAL LATEX CORPORATION 



Girdles 

Brand Awareness 

Purchase Consideration 

Actual Purchase (within four weeks) 

Brassieres 

Brand Awareness 
Purchase Consideration 
Actual Purchase 



Hooper Surveys 








First 

sample 

(1,000) 

Before TV 


Second* 

sample 

(1,000) 

After TV 


Third** 

sample 

(500) 

Before TV 


Third 

sample 

(500) 

After TV 


17.1% 
7.4 
1.4 


20.9% 

11.0 

4.0 


17.4% 
7.8 
1.2 


18.2% 
7.4 
1.2 


7.4 
3.5 
2.4 


7.4 
4.0 
2.4 


8.6 
3.6 
2.2 


11.4 
4.2 
2.4 



♦New sample. 

♦♦Drawn from first sample. 






311 



312 



EXPANDING DEMAND 



1. (a) Can you give me the names of 
several brands of girdles? 

(b) Can you name any others? 

2. The next time you plan to buy a 
girdle, what brands would you con- 
sider buying? 

3. (a) Have you bought any girdles in 
the past four weeks? 

In "no" skip to 4. If yes: 

(b) What brand or brands did you 

buy? 

4. (a) Can you give me the names of 
several brands of brassieres? 

(b) Can you name any others? 

5. The next time you plan to buy a 
brassiere, what brands would you con- 
sider buying? 

6. (a) Have you bought any brassieres 
in the past four weeks? 

In "no" skip to 7. If yes: 

(b) What brand or brands did you 

buy? 

7. Now, I'd like to ask you a few ques- 
tions about your television habits. 
How much time do you spend watch- 
ing TV? From 3 p.m. to 6 p.m.? 
From 6 p.m. to 10 p.m.? After 10 
p.m.? 

8. In the past four weeks, have you seen 
any advertising for girdles on tele- 
vision? What brand was advertised? 

9. In the past four weeks, have you seen 
any advertising for bras on television? 
What brand was advertised? 

10. Is there a television set in working 
order in your home? 



11. Are you employed outside of your 

home? 
Thank you very much. 

The second questionnaire was used on 
the random sample of 500 out of the 
original 1,000. At the insistence of the 
sales department, questions 8 to 10 were 
included in an attempt to correlate 
the shifts in viewers' attitudes with the 
amount of time spent watching TV dur- 
ing the time the spots were shown. The 
research department felt that the respond- 
ents would not answer these last ques- 
tions accurately, but would tend to con- 
nect their favorite brassiere or girdle with 
TV advertising. 

The results of the three Hooper sur- 
veys are shown in Table 4-Z. The attempt 
to correlate shifts in attitudes with the 
amount of time spent watching TV was 
unsuccessful. 



Should the International Latex Cor- 
poration go ahead with its plans for the 
spot television campaign 30 for Playtex 
girdles and brassieres? 



30 Presumably, it would be possible for Latex 
to cancel its advertising contracts at any time 
during the course of the campaign, and thus 
to save 70 per cent to 80 per cent of the ex- 
penditures proposed for the following eight 
months, and 95 per cent to 98 per cent of 
the remaining expenditures. 



EXPANDING DEMAND 



313 



CASE 4-6 

Readiness to Buy 
Certain Consumer Goods 



When a new promotional campaign is 
introduced, no one can breathe easily 
until sales begin to climb and indicate 
success. But must we put up with the 
difficult wait until clear sales evidence is 
available? Are there no reliable indica- 
tors to measure performance before sales 
data come in? 

Even given sales information, what 
about the all-too-frequent and unhappy 
situations when nothing appears to be 
happening or sales seem to be spinning 
downward? Has all the promotion gone 
for nothing; has it contributed nothing to 
the consumer's willingness to buy the 
product? 

Or, on the other hand, has the promo- 
tion just missed the mark, just failed to 
cajole the consumer into a purchase? Could 
some adjustment be made which would re- 
sult in speedier or ultimate sales success? 

That last question in particular is the 
important one that every marketing man 
would like to be able to answer with one 
peek into his crystal ball. 

One way to attack this problem is to 
think of consumers as predisposed (or 
not) to buy your product. That is, how 
ready to buy your product are they? 
Obviously, the consumer who is nearly 
ready to buy is going to be a lot easier to 
sell than his cohort who would not want 
to touch your product under any circum- 
stances. It is the function of advertising to 
create this readiness to buy, and to move 
such predispositions into purchase action. 
Consequently, a reliable measure of con- 
sumers' readiness to buy can have impor- 
tance in evaluating the effectiveness of 



current or proposed marketing strategies, 
through pinpointing the predispositions 
of consumers to purchase a given product, 
thereby highlighting the strengths and 
weaknesses of the campaign. 31 



PSYCHOLOGICAL CONTINUUM 

Naturally, consumers are not equally 
predisposed to buy a given product, serv- 
ice, or brand. As can be seen in Figure 
4-AA, the range of their predispositions 
can be thought of as a psychological con- 
tinuum running from intention to buy 
the item in the immediate future to a 
firm intention not to buy the item ever. 
Between these two extremes, the con- 
tinuum runs through positive intention 
not accompanied by clear-cut buying plans, 
through a neutral area in which disposi- 
tion might swing one way or the other, 
and through an area in which attitudes 
are negative but not firmly set. 

A consumer's position on this con- 
tinuum can be measured by giving the 
respondent a list of brands in a variety 
of product categories, asking him to select 
the statement which most accurately re- 
flects his intentions toward each of them, 
and then expressing the results on an at- 
titude scale similar to the one illustrated 
in the exhibit. This particular scale is 



31 The technique in this case is described by 
William D. Wells, Professor of Psychology 
at Rutgers University and Head of the De- 
velopment Research Unit of Benton & 
Bowles, Inc., which conducted the research. 
See "Measuring Readiness to Buy," Harvard 
Business Review, July- August 1961, p. 81. 



314 



EXPANDING DEMAND 



designed to measure the readiness to buy 
specific brands of frequently purchased 
items. Other versions measure readiness 
to buy large ticket items, infrequently 
purchased items, and services. All scales 
can be adapted by giving appropriate in- 
structions for use in other personal 
interviews or mail questionnaires. 

Other Attitude Measures 

Though measures of readiness to buy 
are closely linked to scales which show 
how well a product is liked, they differ 
from such scales by combining a con- 
sumer's regard for the item with an as- 
sessment of its purchase probability in 
his over-all buying plans. As a result, the 
readiness-to-buy measure is closer to pur- 
chasing behavior, for it taps intention, 
the mental forerunner of action. 

The readiness-to-buy concept also closely 
resembles the concept of "buying inten- 
tions" used in studies conducted by the 
Federal Reserve Board and the Michigan 
Survey Research Center. However, there 



is a major difference in application. Buy- 
ing-intention studies focus on prediction 
of long-term purchasing trends; readiness- 
to-buy studies focus on attitude differences 
for specific products within the present 
market. Consequently, readiness-to-buy 
studies take special note of how far from 
buying the unready consumer is. For the 
marketer, information about readiness not 
to buy can be valuable at times. 

Finally, readiness-to-buy scales are re- 
lated closely to questions like "What 
brand did you buy last?" or "What brand 
do you usually buy?" when the answers 
to such questions will be used to classify 
consumers in evaluating markets. The 
readiness-to-buy measure singles out pros- 
pective customers for special study by 
examining the predispositions of nonpur- 
chasers toward the product in question. 

APPLICATIONS 

Examining consumer predispositions 
makes it possible to think of the market- 



BUYING CONTINUUM 



ATTITUDE SCALE 




Figure 4-AA. Range of Consumers' Predispositions 



EXPANDING DEMAND 



315 



ing process in terms of two broad divi- 
sions. The first, measured by readiness-to- 
buv scales, consists of factors which in- 
fluence the consumer before he reaches 
the point of sale. Thus a consumer's felt 
need for the product or service, his past 
experiences with it, and a variety of other 
influences (including advertising) can all 
be summed up by evaluating his readiness 
to buy the product before he goes out to 
shop. 

The second division contains the condi- 
tions which affect the consumer after he 
has reached the point of purchase, and 
includes such factors as size, location, and 
attractiveness of display; shelf-space allot- 
ment; retailer influence; point-of-purchase 
advertising or price promotion; and the 
influence of other shoppers. 

Naturally, these two divisions of fac- 
tors interact to determine ultimate pur- 
chase action — sometimes reinforcing, and 
sometimes canceling each other. With this 
background in mind, let us examine the 
application of readiness-to-buy scales to 
marketing decisions. 

Trouble Shooting 

By comparing consumers' readiness to 



buy given products with their actual pur- 
chases, it is possible to discover which 
brands are best in converting predisposition 
into action. Such information is important 
because it helps narrow down the search 
for trouble spots when sales are disap- 
pointing. Table 4-BB illustrates the wide 
range which exists among products and 
brands in converting predisposition to 
purchase. But how will such measures help 
pin down trouble spots? 

When sales of a brand are low and 
readiness to buy the brand is high rela- 
tive to other brands in the same product 
category, where is the defect in marketing 
strategy likely to be found? Favorable 
experience with the product and persua- 
sion by advertising are the major ingredi- 
ents of strong predisposition; and when 
such strong predispositions have been cre- 
ated, the advertising and the product itself 
have done as much as can be expected. 
Hence the flaw must be at the point of 
sale. 

However, when readiness to buy is low 
compared with other brands in the same 
category, the marketing deficiency may 
be in the product, or in the advertising, 
or in the interaction between them. 
Brands with this handicap are unlikely to 



TABLE 4-BB. CONVERTING READINESS INTO PURCHASE 
(Base: 900 housewives) 





Toilet goods item 






Grocery item 






Readiness 


Purchasing 


Purchasing/ 




Readiness 


Purchasing 


Purchasing/ 


Brand 


score* 


scored 


readinessX 


Brand 


score* 


scoret 


readinessX 


A 


41% 


23% 


56% 


A 


37% 


23% 


62% 


B 


29 


16 


55 


B 


16 


7 


44 


C 


22 


13 


59 


C 


13 


7 


54 


D 


18 


8 


44 


D 


12 


3 


25 


E 


17 


8 


47 


E 


12 


8 


67 


F 


5 


2 


40 











♦Readiness score: Per cent of respondents who chose one of top two readiness-to-buy scale statements 
about the brand. 

t Purchasing score: Per cent of respondents who actually purchased the brand within four weeks after 
making rating. 

X Purchasing/readiness ratio: Per cent of respondents choosing one of top two statements who purchased 
the brand within four weeks. When this ratio is high, a large proportion of those with intent to buy have actu- 
ally done so. When the ratio is low, something has interfered between predisposition and purchasing. 



316 



EXPANDING DEMAND 



do well; this is true even when conditions 
at the point of sale are excellent. 



Profile Analysis 

Obviously, consumers having different 
demographic or psychological character- 
istics are not equally ready to buy a given 
product. It is therefore frequently impor- 
tant to assess the potential of a given 
product among these different groups. 
Three such assessments can be found in 
Figures 4-CC, 4-DD, and 4-EE. 

In making such comparisons, it is con- 
venient and more meaningful to group 
the readiness-to-buy scale statements into 
larger categories by grouping adjacent 
statements on the scale. These categories 
allow for speedier comparison of the pre- 
disposition to buy for various groups, and 
thereby allow formulation of marketing 
strategy for these groups. 

Category I combines the top two readi- 
ness-to-buy statements ("I am going to 
buy some right away," and "I am going 
to buy some soon"). Consumers who use 
either of these statements are as predis- 



posed to buy as they can be; follow-up 
studies reveal that they actually do buy 
the designated brand in substantially 
greater numbers than do consumers in any 
other category. Again, Category II com- 
bines the next two statements in the scale 
("I am certain I will buy some sometime," 
and "I probably will buy some sometime") . 
The attitude of consumers in this cate- 
gory is one of favorability without im- 
mediate purchasing intention. 

The ratio of Category I to Category 
II is a measure of the degree to which 
favorable disposition has been converted 
into actual purchase intention. A high 
I/II ratio means that many who are 
favorably impressed by the brand actually 
intend to buy it. A low I/II ratio means 
that a "reservoir of good will" remains to 
be exploited; that a large supply of fa- 
vorable disposition still remains to be con- 
verted into firm intent to buy. 

In Figure 4-CC, the low I/II ratio in 
the profile of western consumer opinions 
of Laundry Product X indicates that the 
potential in this region is unexploited. 
Similarly, the high I/II ratios for counties 
with lower population in Figure 4-DD 



/ l 


207o 




1 t] 


397o 




iSillli! 


97o 


n% 




Nt 


m 








nn 



3!7o 



!67o 



(!/n RATIO = 51) 



( I/n RATIO = l.29) 



Figure 4-CC. Regional Profiles with Different l/ll Ratios 



EXPANDING DEMAND 



317 



a 


n 
m 

IV 
Y 


26 % 






357, 




237, 






87, 








87„ 









427c 



277o 



187 



97o 



7 



Figure 4-DD. Population Profiles with Different l/ll Ratios 



57, 



407o 



327o 



207 o 




Note to illustration: County sizes are as 
follows: County size A includes counties 
with cities of over 500,000 population plus 
counties included in metropolitan areas of 
these cities. County size B includes counties 
in which the population is over 100,000 and 
counties included in metropolitan areas of 
cities in these counties (excluding all counties 

indicate that for this Food Product Z, 
the greatest unused opportunity is in the 
highly populated counties. 

A second source of important informa- 
tion in profile analysis is the relationship 
among Categories III, IV, and V. Figure 
4-EE contains profiles with similar pro- 
portions of consumers with immediate 
intention to buy, but with different dis- 
tributions in the remaining profile cate- 
gories. Profiles which differ in this way 
indicate sharply different marketing prob- 
lems. Thus, profiles with the bulge in 
the middle indicate concentrations of 
consumers who have said, "I may buy 
sometime," or "I might buy some some- 



previously defined as "A"). County size C 
includes counties in which the population is 
between 30,000 and 100,000 (excluding all 
counties previously defined as "A" or "B"). 
County size D includes remaining counties 
under 30,000 (not previously defined as "A," 
"B,"or"C"). 



time, but I doubt it." Consumers in this 
frame of mind are still psychologically 
accessible because their intentions are not 
firmly set. 

On the other hand, profiles with the 
bulge at the bottom indicate concen- 
trations of persons who have strong, 
pre-formed negative intentions and are 
therefore likely to be quite difficult to per- 
suade. They may require different appeals; 
or given attempts are likely to be much 
less successful than similar efforts directed 
at consumers in Categories I, II, and III. 
Groups of consumers with such negative 
intentions can often be isolated by demo- 
graphic breakdowns. 



318 



EXPANDING DEMAND 




Figure 4-EE. Profiles with Different Distribution in Categories 111, IV, and V 



NEW BRAND PROGRESS 

If all is going well in a new brand in- 
troduction, predisposition to purchase will 
show a steady increase as the advertising 
begins to take effect and as experience 
with the product confirms established 
expectations. Figure 4-FF illustrates just 
such a happy progression. If the point- 
of-sale activities are also functioning ef- 
fectively, such increased predispositions 
will be converted into rising sales. 

However, if all is not going well, pre- 
disposition measures will help locate the 
trouble. For example: 

( 1 ) Failure to establish ready recogni- 
tion of the brand name appears in pro- 
files as a failure of respondents to move 
out of Category VI ("never heard of the 
brand"). 

(2) Failure to convince consumers that 



the new brand is worth trying appears as 
a movement of respondents from Cate- 
gory VI into Categories III, IV, and V 
instead of into Categories I and II. 

(3) Dissatisfaction with the brand ap- 
pears as a migration into the lower cate- 
gories by those who have already tried it. 

(4) Deficiencies at the point of sale 
appear as a failure to convert predisposi- 
tion into purchases. 

True, not all segments of the consumer 
population respond equally well to initial 
advertising or to initial experience with 
a new product, and different segments of 
the population convert predisposition into 
purchase at different rates. However, seg- 
mentation of the sample will readily show 
which consumer groups are responding 
most favorably, which are converting fa- 
vorable responses into purchases (and 
which are not), which need a little more 
convincing, and which are responding so 



EXPANDING DEMAND 



319 







CIIM 




MON. 




TilF<R 






WFr 


). THUR. 




SAT 








5 






3 






167o 




I 


% 






% 












8% 




87o 




177o 




11 








1 


157o 




■ 


197c 


25 7o 






IV 


51 % 




507o 


317o 




T 


)]% 










187o 




87o 































Tl 



UN. 



107o 



MON, 



TUE?, 



WED. 



THUR. 



2 7o 




Figure 4-FF. Buying Readiness Profiles for Brand X at Six-week Intervals 



unfavorably that they might as well be 
written off, at least for the time being. 



EVALUATING YOUR IMAGE 

Given the many techniques at our dis- 
posal today, it is relatively easy to dis- 
cover clear images for many brands and 
many products. But unless some aspect of 
the image is obviously negative, it is 
difficult to decide what changes should 
be made, if any should be made at all. 
Cross-tabulation of image data with a 
measure of readiness to buy will pin- 
point which ideas are held about a brand 
by those most disposed to buy it, which 
are held by neutrals, and which are held 
by those with negative intentions. 

Figure 4-GG illustrates just such an 
application. In the first graph, the rela- 



tion between image and predisposition is 
such that high readiness to buy is as- 
sociated with the idea that the brand is 
used by intelligent people, while low 
readiness to buy is associated with the 
idea that the brand is used by the unin- 
telligent. 

Unfortunately, not all image elements 
are associated with readiness to buy in 
such a straightforward manner. Thus, in 
the second graph, the association between 
readiness to buy and image variables is 
neither strong nor clear. Being modern is 
presumably a desirable characteristic, and 
it is strongly associated with readiness to 
buy for some products, but not this one. 

The third graph depicts a curvilinear 
relationship. Middle values of this image 
picture are associated with higher readi- 
ness to buy, while lower values are as- 



320 EXPANDING DEMAND 



AVERAGE READINESS-TO-BUY SCORE OF CONSUMER WHO REGARDED "TYPICAL USER" AS 



EXTREMELY VERY FAIRLY SLIGHTLY SLIGHTLY FAIRLY 




I 

n. 

ID. 
IV. 







MODERN 






OLD-FASHIONED 


"■ 




EXTREMELY I 


VERY 


FAIRLY 


SLIGHTLY 


SLIGHTLY 


FAIRLY 


VERY 


EXTREMELY 


I- 
















' -j 


D. 
ID. 


• 


• 


• 


iP"**"*"' 


_^_ 


• 


-^»«M« 


■^^^%^ 


IV. 
V 
















\ 







I^Hi^H^^ 




^dd|B 


~ Rtt&&- 




EXTREMELY 


VERY 


FAiRLY 


SLIGHTLY 


SLIGHTLY 


FAIRLY 


VERY 


EXTREMELY 


I- 


• 


..*-*"" 


• 






""^''"''-•^ 


• \ 
<** 


- 


n 
m- 


• 


****** 


IV- 

V 
















v x 



Figure 4-GG. Three Kinds of Relationship between Readiness to Buy and Image Variables 



EXPANDING DEMAND 



321 



sociated with the extremes of the image. 
In this product category, the most wanted 
brands are thought of as being used by 
people in the middle-age brackets. In this 
way, image elements can be positioned 
in terms of their association with buying 
disposition. A strategy can be subsequently 
developed. 



final outcome. However, they can be 
used to give an advance sign of the suc- 
cess or failure of a marketing plan or to 
"trouble shoot" when sales are not re- 
sponding as expected. In this way the 
marketing strategy, brand image, advertis- 
ing, or other elements of the marketing 
mix can be adjusted or shaped to meet 
the specific market. 



CONCLUSION 

The uses of readiness-to-buy measures 
outlined above suggest that a shift in 
emphasis in the use of predisposition 
measures is needed. So, instead of such 
measures being viewed solely as predictors 
of behavior, they should be viewed both 
as predictors and as market data worthy 
of independent consideration. This is not 
to suggest that predisposition measures 
can replace sales figures as evidence of 



How many kinds of marketing deci- 
sions would be improved by the use of 
this technique? Can it be applied to in- 
dustrial products? What does it contribute 
to understanding how advertising works? 
To understanding consumer behavior? 
Does it have any significance for the pre- 
vious cases in this chapter, 4-1 to 4-5, or 
for earlier cases like 2-1, 2-3, 2-4, 3-6, 
3-10? 



322 



EXPANDING DEMAND 



CASE 4-7 

Tawland Corporation 



A Tawland research and development 
project started in 1958 and successfully 
completed in early 1960, resulted in the 
"Trackmobile," a small gasoline motor- 
driven vehicle which had both rubber 
tires enabling it to travel on pavements 
and, at right angles, a set of standard 
railroad wheels so that it could travel 
on United States gauge railroad tracks. 
The rubber-tired wheels could be hy- 
draulically lifted making it possible for 
the vehicles to cross a set of tracks, settle 
down upon them, and continue travel 
on the tracks. In the opinion of execu- 
tives, the Trackmobile would replace 
many switching locomotives in use in 
many factory yards. Although it was 
claimed that the vehicle could pull up to 
200 tons in actual tests, the Trackmobile 
had moved up to 550 tons of loaded 
freight cars. Traction to move such loads 
was obtained by means of a hydraulic 
jack which slightly lifted the first at- 
tached freight car, thereby shifting the 
load to the wheels of the Trackmobile. It 
was powered by a 60-horsepower jeep en- 
gine which was sufficient to equal the 
performance of 20-ton switch locomotives. 
Patents had been applied for, but by 
January, 1961, these had not been granted. 

Several factors indicated to executives 
that a lucrative market existed. While 
switch locomotives cost $15,000 to $20,- 
000, and diesel locomotives cost from 
$20,000 to $75,000, the Trackmobile, 
which, in many cases, served the same 
purpose and was much more versatile, 
had much lower capital cost. Another 
appeal was its weight 32 of only 6,000 

32 The Trackmobile was light enough to be 
transported by plane. 



pounds as contrasted to a locomotive 
which weighed 100,000 pounds. In view 
of this weight factor, executives believed 
the company should have no difficulty 
in obtaining steel to manufacture it, 
since the government should be glad to 
give priorities to an item which actually 
saved steel. 

The executives of Tawland Corpora-, 
tion were particularly interested in the 
aggressive promotion of the Trackmobile, 
since it appeared to them that such an 
item should be less subject to cyclical 
fluctuations in demand than the com- 
pany's other products. Because executives 
decided that they did not want to dis- 
tribute this product directly to the final 
user, it was evident that some type of 
distributor was needed. It was finally de- 
cided that the materials handling spe- 
cialist would be the most suitable. These 
specialists were firms which sold relatively 
few products, specializing in such items as 
lift trucks, small tractors, and similar 
products. 

Originally, it had been suggested that 
the best way of introducing the Track- 
mobile would be to build a special trailer 
for it and sell it throughout the country 
to distributors. However, the vice presi- 
dent in charge of the wholesale division 
believed this type of introduction would 
not be sufficient. He felt it took show- 
manship to sell distributors and believed 
that the introductory program must be 
carefully planned and timed, and that 
its focus should be a show which was to 
include a striking demonstration. 

Accordingly, on November 15, 1960, a 
large show was held in Grand Cen- 
tral Station in Chicago, Illinois. To this 



EXPANDING DEMAND 



323 



show, 300 members of the press, 100 
leading industrialists, and 15 to 20 bank- 
ers had been invited. It had been ar- 
ranged to use the station during a time 
that no trains were scheduled for arrival 
or departure. Facilities for cocktails, a 
loud-speaker system, guides, and other 
features believed necessary for a successful 
show were included. The Baltimore & 
Ohio Railroad had made available to the 
company loaded and unloaded freight 
cars, coal cars, and switch locomotives. 
Guests were first invited by mail. If they 
did not reply, a telegram was sent. If 
they still had not answered, they were 
called by telephone. This assured the com- 
pany that attendance would be good. In 
order to make sure that a dealer organiza- 
tion would exist to back up the show, 
executives of Whiting had secretly shown 
the Trackmobile to 25 distributors and 
had given them franchises. 

The official presentation of the Track- 
mobile was later described by the vice 
president as a "phenomenal" success. The 
company received free publicity releases 
in Time, Fortune, Business Week, Look, 
and several other leading magazines, stories 
in leading newspapers including the St. 
Louis Post Dispatch; and articles in most 
trade journals. The show, which cost 
$5,000 to $6,000, yielded an advertising 
return considered to be worth many times 
this figure. On January 24, 1961, the 
company was still receiving inquiries at 
the rate of 12 per day, although no ad- 
vertising expenditures had been made. 
While not one Trackmobile had as yet 
been produced, the company had $450,- 
000 worth of orders on its books. 



Tawland executives had come to the 
conclusion that the distributors should 
sell the product on the basis of demon- 
strations in the customers' plants, be- 
cause experience had indicated that four 
out of five demonstrations resulted in a 
sale. In order to assure the proper utiliza- 
tion of the Trackmobiles, the conditions 
of sale to a user included one clause 
permitting the distributor to spend up to 
three days in a customer's plant to in- 
struct the latter's personnel, and another 
clause stating that the distributor alone 
was to be allowed to furnish technical 
service. 

The price of a Trackmobile was $7,950 
delivered. Distributors paid $6,375 f.o.b. 
plant, Harvey, Illinois. It was carefully 
pointed out to the distributors, however, 
that their margin included a $100 provi- 
sion for the cost of demonstration, $175 
freight, $100 cost of instruction, and 
$75 for three months' service, making 
the dealer's full cost $6,825. 

At the end of January, 1961, the Taw- 
land Corporation was setting up produc- 
tion for the Trackmobile. In view of the 
great success of the Chicago show, a simi- 
lar introductory show was held in Mon- 
treal in the first week of February, in 
order to introduce the Trackmobile in 
Canada. 



Is this a well-conceived promotional 
program? Can Tawland look forward to 
a good volume of sales? What is involved 
in building demand for a product like 
the Trackmobile? 



324 



EXPANDING DEMAND 



CASE 4-8 

Whiting Corporation 



The principal product line of the whole- 
sale division of the Whiting Corporation 
was electric chain hoists, small electrically 
operated mechanical hoists which had 
numerous applications throughout indus- 
try. Up to April, 1960, executives of the 
Whiting Corporation lacked confidence 
in the company's ability to distribute the 
product, since the Whiting name was 
relatively unknown in the chain hoist 
business. Executives were inclined to con- 
clude that they could not get the leading 
mill supply houses to handle the product. 
Furthermore, executives believed that, in 
view of these circumstances, the Whiting 
product had to be sold below the es- 
tablished market price. The hoists, which 
were quoted at an average price of $230 
a unit, were generally sold to the final 
user by the mill supply house salesman 
who called upon the customer together 
with Whiting representatives. These cus- 
tomers would sometimes be entertained, 
and extensive negotiations would often 
follow. In April, 1960, the new vice presi- 
dent of the wholesale division became 
convinced that this type of distribution 
was both ineffective and unprofitable, 
since selling expenses exceeded $100 per 
hoist while the company's gross margin 
on the product was only $60. 

One of the first steps undertaken by 
the vice president of the wholesale di- 
vision was to rewrite the catalogue de- 
scribing the electric chain hoist in order 
that it have more appeal to the ultimate 
buyer and the distributor. The new cata- 
logue emphasized the usefulness of elec- 
tric chain hoists; while the previous cata- 
logue had concerned itself primarily with 



explaining the technical features of the 
Whiting product. After the new catalogue 
was completed, the wholesale division 
embarked upon what it termed the "cream 
of the crop" program. 

The initial step in the plan was to 
establish a strong distributor organization. 
Executives determined the number and 
type of distributors that were desired in 
each area. Each salesman was then in- 
structed to call upon the best distributors 
in his territory in an effort to get them 
to handle the hoists. In order to assure 
the selection of good distributors which 
would aggressively sell the product, the 
requirement was set that each distributor 
must maintain at least five electric hoists 
in stock and have at least three salesmen 
in his employ. 

An adjunct to the "cream of the crop" 
program was a new discount schedule. 
The standard trade discount given by 
manufacturers to distributors was 25 per 
cent and 5 per cent. However, in launch- 
ing the "cream of the crop" plan, addi- 
tional discounts for volume were to be 
given. If a distributor, during the first 
six months of the fiscal year, bought eight 
hoists, he obtained a retroactive discount 
of 3 per cent. For ten hoists he obtained 
4 per cent, 15 hoists 5 per cent, 22 hoists 
7 per cent, and 30 hoists 10 per cent. 
These discounts were given at the end 
of each six months' period starting Sep- 
tember 1, 1950. It also was necessary for 
salesmen to convince existing distributors 
to cooperate with the company in its 
"cream of the crop" plan. By January, 
1951, 115 distributors were selling the 
company's electric chain hoists, and the 



EXPANDING DEMAND 



325 



vice president hoped to increase this num- 
ber to 150. 

Once the manager of the distributing 
organization had agreed to cooperate in 
the "cream of the crop" plan, a sales 
meeting was arranged for the distribu- 
tor's salesmen, and the Whiting salesman 
presented the plan to the salesforce. As 
an aid in making this presentation, each 
Whiting salesman had in his possession a 
presentation manual which had cost $500 
to prepare, which he would place in front 
of the sales meeting, and by turning the 
pages of this manual make his presenta- 
tion. 

As part of the plan, the distributors' 
salesmen were supplied with a smaller 
presentation manual which they were to 
use in selling the hoist to the final user. 
In order to make sure that the distribu- 
tors' salesmen would actually make use 
of the books, each distributor was limited 
to ordering only a sufficient number to 
supply to salesmen selling hoists. Further- 
more, he was sent an invoice memorandum 
charging him $5 per book. This invoice 
was not to be paid, but it was intended 
to serve as a reminder that the books had 
a value and remained the propertv of the 
Whiting Corporation. In the books them- 
selves, one sales feature of the Whiting 
hoist was described on each page in lan- 
guage addressed to the distributor sales- 
men. Thus, the salesmen using the book 
would point out to the customer that 
the manual was intended only for sales- 
men. The vice president believed that a 
distributor salesman, using the sales man- 
ual as a selling aid, would be forced to 
give a slower and more complete and 
understandable presentation of the prod- 
uct. A supply of small giveaway catalogues 
describing the hoists was furnished to 
each distributor. 

During the presentation to the distribu- 
tor's sales force, the Whiting salesmen 
stressed the value of selling a Whiting 



product, pointing out the size and repu- 
tation of the company and the wide range 
of products it manufactured. The presen- 
tation then entered the area of pointing 
out why hoists were needed and why a 
distributor should sell hoists; and lastly, 
point by point, the superiority of Whit- 
ing hoists. Since this presentation paral- 
leled the copy in the small giveaway 
catalogue, it was hoped that the distribu- 
tors' salesmen, when explaining the com- 
pany's catalogue to a prospect, would give 
a similar detailed presentation. After the 
Whiting salesman had explained the elec- 
tric chain hoist, he proceeded to present 
the "cream of the crop" plan. 

The basis of the plan was a belief that 
20 per cent of its customers accounted 
for 80 per cent of the business for each 
distributor. The "cream of the crop" plan 
was designed to make the distributors' 
salesmen concentrate their efforts upon 
this 20 per cent. The Whiting plan in- 
cluded active support of such efforts with 
a direct-mail advertising campaign. 

In order to determine which firms were 
the "cream of the crop," each distributor 
salesman was asked to list those firms in 
which lifting operations were performed. 
After such a preliminary study, he was 
directed to compile a list of 30 accounts 
which he wrote down on a special form 
(the first page of which is illustrated in 
Figure 4-HH and 4-II). (It was estimated 
that each distributor salesman had 150 
active accounts.) On this form, in addi- 
tion to naming the firm and the type of 
operation, he specifically mentioned the 
individual or individuals who were re- 
sponsible for buying or approving the 
purchase of the hoist. These complicated 
forms were then given to the advertising 
agency of the Whiting Corporation which 
subsequently sent three mailings to each 
name on the distributors' stationery over 
the salesman's signature. 

The first mailing was a short, person- 



Figure 4-HH 




WHITING HOIST 




Here's the first step to more sales of Whiting Elec- 
tric Hoists! Each prospect name you list inside will 
receive a series of 3 specially personalized mailings 
... on your letterhead . . . signed by you . . . 
featuring the money-saving advantages of Whiting 
Electric Hoists. 
PIEASE REMEMBER . . . LIST ONLY YOUR VERY BEST PROSPECTS! 

1. Fill out completely all personal names, company names, 
addresses, etc. 

2. For large firms, list the names of such individuals as president, 
works manager, production manager and purchasing agent or 
whoever is responsible or influential in the purchase of Whiting 
Electric Hoists. For small shops, list both the owner or mana- 
ger and the chief mechanic. 

3. For each personal name you list, please give us three of 
your company's letterheads and envelopes plus 15 extra for 
possible spoilage. 

THESE MAILINGS WILL REALLY SOFTEN YOUR PROSPECTS . . . 

THEY WILL BE READY FOR YOUR CALLI 

FILL OUT YOUR LIST NOW) 

WHITING CORPORATION 

HARVEY, ILLINOIS 



EXPANDING DEMAND 



327 



tftfUITINIt DISTRIBUTOR SALES MANAGER'S COPY 

" ■■■ ™ 11^ ^P This is your copy of the prospects selectee 

for the Whiting Hoist "Cream-of-the-Crop" 

ELECTRIC plan by the salesman whose name is 

shown at right. He has also been sent a 

M^%| CT c °py °f Tn ' s which carried the following 
■■^PBwB suggestions: 

1. Keep this list with you constantly and use it 

in planning your calls to make certain that 

each prospect is followed up as- soon as 

practical. 

frffrflX^Jf fiflrfnTP flfiftti ^" Note ,ne hiailing schedule and tie-in your 

LAJUflfJ- VI iP-Jj yfiyT 3. Use* co'de letters to keep record of calls and 




List No. consisting of 

names 

from 

(names and initials of Distributor Salesman) 
(name of Distributor Company) 


4. Advise your Sales Manager regarding results 
of your calls after follow-up of 3rd mailing. 

D| AM WE SUGGEST THAT YOU CHECK EACH 
rLAN SALESMAN'S LIST REGULARLY. THIS 

WILL HELP ASSURE THE SUCCESS OF THE 

PLAN! 


MAILING SCHEDULE 
Letter and Enclosure No. 1 










CODE KEYS: 




SOUWX" |CAllBACK-"Y" PLUS DATE 






(PLACE BENEATH NUMBERS) 


NEEDS MORE PROMOTION-"Z"| REMOVE FROM LIST "O" 


Letter and Enclosure No. 3 













ALL NAMES BELOW ARE FOR 



UNLESS OTHERWISE SPECIFIED 



City 



I. 


2. 


3. 


4. 


5. 


6. 



Figure 4-11 



alized letter with an enclosure (Figure 
4-JJ). The second mailing was sent two 
weeks later, followed by the third mailing 
in two weeks, each consisting of a short 
letter and an enclosure. The Whiting 
Corporation expected to receive about 
10,000 names, and it was estimated that 
the cost of the campaign would be 6 
cents per name per mailing. A list of the 
prospects, the dates of the mailing, plus 



a copy of the mailing was sent to each 
distributor salesman, who was expected 
to call upon the accounts which received 
the mailings. 

In order to check up on the effective- 
ness of the campaign, the Whiting Cor- 
poration required that the distributor sales 
manager, together with the salesman, 
check the results by listing the actions 
taken by each account on another special 



328 EXPANDING DEMAND 



Some UKiw| Methods 

AM TNI MAK OP IN-IMCIM M€YI 




Why not .. 
lift the faster, 
safer, less costly v 
way with a... ; 

WHITING 

ELECTRIC HOIST 



WHITING CORPORATION, Harvey, Illinois 

Manufacturer of Cranes, Hoists and other 
Equipment for over 60 years. 



Figure 4-JJ 



EXPANDING DEMAND 



329 



form. On this form, indication was given 
as to whether the account was sold, 
whether the salesman was requested to 
call back, whether additional promotions 
would be required, or whether it should 
be dropped from the list. The forms were 
also checked by the Whiting Corpora- 
tion salesmen, Whiting Corporation ex- 
ecutives, and the advertising agency. 

Another advantage of this plan, ac- 
cording to the vice president, was that it 
was possible for the advertising agency to 
tabulate the list of 10,000 names in order 
to determine the positions of individuals 
who were generally responsible for the 
purchase of a hoist. This information was 
then to be used in selecting the media 
of the company's national advertising 
campaign. 

The plan was presented to the com- 
pany's salesmen in August, 1960, and was 
first explained to the districts in Novem- 
ber, with the first mailing during that 
month. Executives believed that this type 
of campaign would stimulate the dis- 
tributor salesmen to sell a considerable 
number of hoists, and establish the Whit- 
ing name in the hoist business, where the 
company had been relatively unknown. 

In the spring of 1961, executives pointed 
out that a second "cream of the crop" 
campaign would probably not be under- 



taken in the foreseeable future in view 
of the company's oversold condition. 
However, no distributor was to be deprived 
of the opportunity of having an initial 
campaign in his territory. 

When, in the future, the aggressive 
selling of chain hoists was to become neces- 
sary once more, it was expected that the 
distributors would share some portion of 
the cost of the campaign. Executives 
pointed out that, when the full results of 
the 1960 campaign were measured, they 
could then decide whether the 30-name 
limit placed upon each distributors' sales- 
man should be lowered or raised. If the 
plan proved successful, executives believed 
that it should be applied to some of the 
company's other wholesale products, par- 
ticularly the rotary shears and trimmers 
which were sold through machinery sup- 
ply houses. 



Is this a well-conceived promotional 
program? How does it compare with 4-7 
for effectiveness in building sales of the 
particular product? Where and how does 
the buying urge get put into motion in a 
situation like this one? Does Whiting 
have the right kind of distribution setup? 



330 



EXPANDING DEMAND 



CASE 4-9 

Lightning Aircraft Company 



In 1959, the Lightning Aircraft Com- 
pany, which manufactured single-engine 
planes, bought for purposes of sport and 
business use, was considering expanding 
its line to capture a larger share of the 
latter market. More and more corpora- 
tions were buying planes to transport 
their top executives on business trips, and 
the Lightning management was attracted 
both by the number of possible buyers 
and by the size of the unit sale (since 
business planes tended to be larger than 
the existing Lightning models). 

Lists of present corporate owners of 
planes were found, and questionnaires were 
mailed to a stratified sample, as follows: 

Actual number 
of companies 
owning planes 



Single engine 
Light twin 
Heavy twin 



12,760 

2,083 

450 

15,293 



Cross-checking of respondents with the 
original lists and follow-up of nonrespond- 
ents seemed to indicate that returns were 
representative geographically, by size of 
company, and so forth. About two thirds 
of the returns came from company presi- 
dents (to whom the questionnaires had 
originally been addressed, except in the 
case of heavy twin-engine owners, mostly 
large companies, where about half of the 
returns were filled out by the aviation 
department or chief pilot). 

Because management considered three 
variables important — speed, capacity, and 
price — the principal question provided a 
matrix with 64 possible combinations of 



the three factors, and respondents were 
asked to enter in the appropriate box or 
boxes the number of planes of that type 
which they expected to purchase over the 
next three years. As shown in Figure 
4-KK, the 64 combinations were then 
divided into six classifications as follows: 

I. Single-engine planes 
II. Light twin-engines 

III. Heavy twin-engines 

IV. Turboprops 
V. Turbojets 

VI. All others (mostly existing military 
or commercial airline planes con- 
verted to private use) 



Names avail- 
able for 
mailing 

640 

1,689 

450 

2,779 



Number of 
returns 

183 
530 
173 

896 



Further, returns were analyzed accord- 
ing to present ownership of planes: (a) 
single-engine, (b) light twin-engine, and 
(c) heavy twin-engine. Turboprop and 
turbo-jet planes were not being offered to 
the corporate market at this time. The 
results came out as follows: 



/// IV 



VI All 



a 


61 


30 


10 


- 


- 


- 


101 


b 


52 


80 


158 


17 


3 


46 


356 


c 


2 


2 


19 


14 


16 


30 


83 



Total 115 112 187 31 19 76 540 

Analysis of present owners by brand 
showed no meaningful relationships. That 
is, a present Lightning owner was no more 



o 
o 

© 

CD 
CM 


O 
O 

o 

8 

CO 


O 
O 

o 
o 
in 


o 
o 
o 

o 

o 
t- 


© 

8 

©~ 

o 
o 

i-H 

^5- 


o 

8 

o 
o 

CM 

i-H 


o 
o 
o 

o 

o 
co 

i-H 


o 
o 
o 

o 

o 
o 
co" 


O 

o 
©^ 
©~ 

CO 
CM 

€^ 


O 

o 
<=l 

o 

CD 
CM 


o 
o 
o 

o 

o 


© 
o 
©^ 

o" 
o 

in 


o 

o 

o 
o 

CO 


o 
o 

©~ 
o 

1-1 


o 
o 
o 

o 
© 
m 

t-H 


o 
o 
o 

o~ 
o 

CO 

cm" 


O 

o 

CO 

©~ 
o 

CM 
€«- 


8 

in 

CM 
CM 


O 

o 

o~ 
o 

CO 


o 
o 
©^ 
in" 

CO 


o 
o 

o" 
o 
in 


O 
O 
O 

©" 

O 
CO 


O 
O 
©^ 

o" 

O 
O 


o 
o 

o" 

o 

CD 

i-H 


o 
o 
©^ 

o 
m 


O 

o 

©^ 
in 
c- 

*— i 


o 
o 
©^ 

©~ 

CM 
CM 


© 
o 

©" 

00 
CM 


o 
o 

o" 
o 


© 
o 
©^ 

o" 
o 

CD 


o 
o 
o 

©" 
o 

co 


o 
o 
o 

o" 
o 

1—1 


O 

o 
o 

€»- 


O 

O 

©~ 

CM 


© 
o 
o 
1ft 


o 
o 

©" 

CM 


o 
© 
o 

o" 

CD 
CM 

€«- 


O 

o 
o 

©" 
o 


o 
o 
o 

o" 
o 
c- 

€«- 


o 
o 
o 

o" 
o 

CO 
^3- 


o 
© 

CO 

in 

CD 
€^ 


O 
O 


O 
O 

o" 

CM 

i— ( 


o 
© 

CD 

in" 

CD 


O 
O 
O 

©" 

O 
CM 


O 
O 

©" 
O 

CO 
€»- 


O 
O 
©^ 

©" 
o 

ID 


o 
o 

o 

o" 
o 

CD 


O 

o 

CO 

in 

CM 


O 

o 

©~ 
m 


o 
o 
©^ 
in 


O 

o 

C3 

©" 

O 

l-H 


O 

o 
©" 

CO 


O 
O 
O 

©" 
CM 
CM 


o 
o 
o 

o 
m 

CO 


o 
o 
o 

o 
in 
■<* 


o 
o 

CD 
CN 


o 
o 
o 

CD~ 
i— t 


o 
o 
o 
o" 


© 
o 
©^ 
o" 

00 


o 
o 
o 

o 


o 
o 

o" 

CO 
55- 


O 
O 
O 

O 

m 

CM 


O 

o 
o 
o 

CD 
CO 
€«- 






^a. 



331 



332 



EXPANDING DEMAND 



or no less likely to buy a Class II plane 
than owners of other, competing Class 
I planes. 

It would cost Lighting about three- 
quarters of a million dollars to develop 
and tool for a Class II or Class II model, 
and at least several million dollars for a 
Class V or V model. The company 
figured that manufacturing, selling, and 
administrative costs would be approxi- 
mately 90 per cent of selling price, and 
that it would need to recover development 
costs in three years. The company was 
currently enjoying a market share of 9 
per cent in the single-engine field, but 
management realized it would take two 
years to get to the market with twin- 
engines and then five or six years before 
it could make as much headway in the 
twin-engine field as in the single-engine 
field. The turboprop and jet market was 
even farther away because it would re- 
quire considerable change-over in the 
company's operations, but management 
still felt it should try to gauge future 
potentials fully. 

An additional mailing to a random 
sample of 757, with 203 returns, uncov- 
ered 17 purchase indications on the part 



of present nonowners of planes, over half 
of them qualified by "maybe." Of the 17, 
the majority (11) favored Class I planes. 
The names of this mailing were taken 
from a list of 70,000 firms which were 
believed to be large enough to afford and 
find a company plane useful. The num- 
bers of planes delivered for other than 
military or commercial airline use had 
been 6,413 in 1958, with twins account- 
ing for 894 of these. 33 As of January 1, 
1958, the business aircraft fleet was re- 
ported to be composed as follows: 34 



Single engine 
Twin- engine 
Miscellaneous 



24,470 

3,842 

171 

28,483 



Can the results of the survey be pro- 
jected to the total market? Will the re- 
sults hold over time? What class of plane, 
II, III, IV, or V, is likely to be most 
profitable for Lightning's expansion move? 



33 Aviation Week, March 9, 1959. 
™Ibid., March 9, 1958. 



EXPANDING DEMAND 



333 



CASE 4-1 

Analysis of Distribution Costs 



It is difficult to exaggerate the oppor- 
tunities for reduced marketing costs and 
increased marketing efficiency, and hence, 
greater profits, which are offered to man- 
agement by the combined techniques of 
distribution cost analysis and mathematical 
programing. In the offing may be a revo- 
lution in the planning and execution of 
distribution that is fully comparable to the 
triumph of time and motion studies and 
cost analysis in the factory. 

In the pages to follow we shall de- 
scribe 35 some of the more important tech- 
niques that management can profitably 
use, and we shall show that: 

(1) The misallocation of marketing 
effort in industry is greater than most 
people realize. 

(2) The separation of (a) fixed costs 
incurred in common for different types of 
sales effort, (b) separable fixed costs, and 
(c) variable costs which are related to 
different segments of the business is one 
of the steps in analyzing a company's dis- 
tribution problem. 

(3) The techniques of mathematical 
and linear programing can be used to de- 
rive approximately, on the basis of these 
cost data, the most profitable allocation 
of a firm's marketing expenditures and re- 
sources. 

(4) Once the needed data on cost- 
volume relationships are obtained and 



35 This approach is described by William J. 
Baumol, Professor of Economics, Princeton 
University, and Charles H. Sevin, Senior As- 
sociate, Alderson Associates. See "Marketing 
Costs and Mathematical Programing," Har- 
vard Business Review, September-October 
1957, p. 52. 



the computations are completed, the 
businessman can proceed to redirect his 
marketing effort in a way that is virtually 
guaranteed to increase his profits. 

In a short appendix we shall suggest a 
few inexpensive computations that will 
help to improve distribution cost analysis. 



TREMENDOUS OPPORTUNITIES 

In most businesses a very large propor- 
tion of the customers, orders, products, 
and territories brings in only a very small 
proportion of the sales. But selling, ad- 
vertising, and other marketing efforts all 
too frequently are expended in proportion 
to the area covered, the number of cus- 
tomers, the number of orders, and so 
forth; management does not give enough 
explicit consideration to their actual and i 
potential contribution to sales volume/ 
and profit. 

Even the better-managed firms seldom 
realize how much of their marketing ef- 
fort brings in only very small sales re- ' 
turns, since it is difficult to find out which 
sales can be ascribed to which selling and 
promotion work. They make little or no 
systematic attempt to evaluate the results 
of specific portions of marketing effort, 
and usually measure their success solely 
by the firm's total dollar sales in each 
product line. Moreover, the manufactur- 
ers of branded consumer goods with a 
national market typically follow a policy 
of 100 per cent coverage of the market 
in order to support their national ad- 
vertising. 

For these and other reasons, there is 



334 



EXPANDING DEMAND 



widespread misallocation of sales effort. 
A business as a whole may be making a 
good profit; but if you analyze its costs 
and sales carefully, you will usually find 
that a large number of sales are not very 
profitable at all — at least, as compared 
with certain other sales. 

Also, such sales are a heavy drain on 
potential profits. When time on an expen- 
sive television program is devoted to the 
promotion of a low-profit item, or a sales- 
man spends time on an unpromising re- 
tailer, or limited warehouse space is tied 
up by large stocks of a low-turnover, low- 
markup item, the costs to the firm are 
high. Valuable television time, sales ef- 
fort, and warehouse space are thereby 
withheld from more profitable uses. If a 
salesman divides his time between one 
product which earns the company $5 
an hour and another which nets $12 an 
hour, then every hour spent promoting 
the former, in effect, costs the company 
$7. 

This is quite elementary, yet its lesson 
is ignored all about us every day. For 
instance, one company made a distribu- 
tion cost analysis and found that 68 per 
cent of its customers, bringing in only 10 
per cent of the volume, were responsible 
for a net loss of as much as 44 per cent 
of sales. A distribution cost analysis in 
another firm revealed that 95 per cent of 
all the customers in one territory were 
unprofitable — with losses ranging up to 
86 per cent of sales. 



Profit Tools 

The substantial losses on unprofitable 
sales resulting from the disproportionate 
spreading of marketing effort can be 
minimized or even eliminated, simply by 
making certain that the marketing dollar 
goes where it can do the most good. This 
can be done with the help of two relat- 



ed tools — distribution cost analysis and 
mathematical programing. These tools can 
indicate to management where and how 
to apportion marketing effort to make the 
most of potential net profit possibilities. 
In fact, companies which have used just 
distribution cost analysis as a management 
tool — without the use of mathematical 
programing — have achieved startling re- 
ductions in their distribution costs by 
correcting only the more obvious mal- 
distributions of market effort. 

For example, in one company market- 
ing expenses were cut nearly in half, from 
22.8 per cent to 11.5 per cent of sales, 
and a net loss of 2.9 per cent on the books 
was turned into a net profit of 15 per 
cent, after shifting some effort from the 
68 per cent of accounts which had been 
unprofitable. Another company shifted 
selling and advertising effort from less 
profitable to more profitable territories 
and achieved a 78 per cent increase in 
average sales per salesman, a reduction 
of 33 per cent in the ratio of selling and 
advertising expense to sales, and an in- 
crease of about 100 per cent in the ratio 
of net profits to sales. 36 



Steps in Cost Analysis 

How is management to find out what 
the different parts of the firm's marketing 
process contribute to its costs, its profits, 
and its sales? It is not so easy as it sounds. 
Prevalent accounting techniques for re- 
cording the results of marketing activities 
are insufficiently detailed; their informa- 
tion is distorted by arbitrary cost alloca- 
tions, and their figures are only part of 
what is required. 

The first step is a finer breakdown of 
the firm's average cost and profit data. 



36 These two cases are taken from Charles H. 
Sevin, How Manufacturers Reduce Their 
Distribution Costs (Washington, Govern- 
ment Printing Office, 1948) . 



EXPANDING DEMAND 



335 



The over-all distribution costs for the en- 
tire business must be allocated to the 
specific segments of the business for which 
they are incurred. For example, through 
distribution cost analysis we find that the 
sale of a thousand cases of Product A 
through medium-size retailers located in 
the Chicago metropolitan area requires 
x dollars worth of salesman time, y dol- 
lars in transportation and warehousing 
costs, z dollars in advertising expenditure, 
and so on. We then get the production 
costs and figure the net profits or losses 
for each segment separately. This is not 
the place for an extended discussion of 
the principles and methods of distribution 
cost analysis, but two basic principles of 
the techniques used can readily be sum- 
marized: 

(1) The distribution expenditures of a 
particular business, which are usually re- 
corded on a natural-expense basis, are 
re-classified into functional-cost groups, 
which bring together all of the indirect 
costs associated with each marketing ac- 
tivity or function performed by that com- 
pany. 

( 2 ) The functional-cost groups are allo- 
cated to products, customers, and other 
segments of sales on the basis of meas- 
urable factors, or product and customer 
characteristics which bear a cause-and- 
effect relationship to the total amounts 
of these functional costs. 37 



TREATMENT OF COSTS 

Fixed (overhead) costs may be defined 
as those which, in the short run, do not 
change in total amount as sales increase 
(although, as we shall see later, costs 
which are fixed in one problem can be 



37 For further details on classification of dis- 
tribution costs by functional categories and 
bases for their allocation to specific segments 
of a business, see Charles H. Sevin, op. cit. 



variable in another). This may be the 
result of contractual obligations assumed 
by the firm, or it may represent salaries, 
sunk or irrecoverable expenditures in 
buildings, equipment, and so forth. 

Not all fixed costs should be treated 
alike, however. In fact, one of the crucial 
steps in distribution cost analysis is to 
distinguish between those costs which can 
be charged to specific types of sales and 
those which cannot. The management 
which is ready and able to do this well 
can confidently go on with the job of 
deciding where its unrealized potentials 
for profit lie. 

Exclude If Common 

Some fixed marketing costs are incurred 
in common for several different sales seg- 
ments — for example, the advertising of a 
company's brand name, which must auto- 
matically influence the sale of all package 
sizes in one degree or another. Any break- 
down of these advertising costs among the 
various package sizes must be arbitrary 
because the facts only entitle us to say 
that the advertising program is serving all 
of the package sizes at once. 

Here we run into a serious problem. 
Two different methods of allocating such 
fixed costs can easily yield totally differ- 
ent results, and clearly both cannot be 
right. There is, in fact, no really correct 
method of allocating fixed costs incurred 
in common by sales for different products 
or customers. 

The solution to this problem is simple: 
fixed costs which are common to several 
sales segments and whose magnitudes do 
not vary with the volume of sales in any 
one segment should be omitted from all 
cost and profit computations. Sales in- 
come should be viewed as a contribution 
to profit and fixed cost together. The 
situation is like a family debt; what each 



336 



EXPANDING DEMAND 



brother brings in helps both to support 
the family and to pay off the family's 
obligations, although it does not affect 
the size of the rent, food, and heating 
bills. 



Include If Variable 

None of this applies to variable costs, 
which increase with sales and therefore 
can be allocated. For instance, more sales 
in Wisconsin help to pay the company's 
transportation bill, but they also increase 
the magnitude of this bill. It is a variable 
cost. The addition to the freight bill 
cannot be ignored in seeing how much 
these Wisconsin sales are worth to the 
firm. But a fixed cost of the sort we are 
discussing does not vary with sales. The 
added income from Wisconsin contrib- 
utes to the payment of fixed costs but 
adds nothing to their size. 

It follows, therefore, that an allocation 
of marketing costs which omits fixed ele- 
ments can show how much a particular 
type of sale contributes to the well-being 
of the firm and can indicate how a re- 
vision of the firm's distribution efforts 
can help sales and profits. Costs which 
include arbitrarily allocated fixed elements 
can be seriously misleading, and they do 
not add any useful knowledge to a man- 
agement analysis of sales and profits. 



PROFIT 



$5,000 


1 \. 
1 \ 
1 \ 
1 \ 


...i--- 


' 1 *\ N. 

I x ^"■—--- - 

i 




i 

i 

I *^„ 

l ^~-- 

i 

i 
i 



SALES 
Figure 4-LL 



Suppose that a company correlates its 
sales and profits in a certain territory and 
finds a curve like the solid line in Figure 
4-LL. It learns that the sales dollar is 
most profitable when total sales volume 
is at point x. Now suppose that ware- 
house rental goes up and adds $5,000 to 
fixed cost. This lowers the profit curve at 
every point; the dotted line now repre- 
sents profits. But the highest point of the 
new curve still occurs at x, simply because 
warehouse rental is a fixed cost at all 
sales volumes. A change in such a cost 
does not tell management anything it did 
not know before about the sales-maximum 
profits relationship. It would be impor- 
tant to know only in the exceptional case 
where the increased fixed costs make it 
impossible for the firm to break even. 

The elimination of fixed expenses from 
the allocation of costs is not quite so 
simple in practice as it may sound. It is 
often difficult to distinguish between vari- 
able and fixed marketing costs. For in- 
stance: 

(1) Warehousing cost, which is fixed 
when the warehouse is not used to capac- 
ity, becomes variable when, with all the 
storage space filled up, management con- 
siders the construction of more space to 
eliminate a bottleneck. 

(2) A sales manager's salary may be 
a fixed cost unless the company considers 
firing him altogether rather than trans- 
ferring his services from one product line 
to another. 

(3) Most, if not all, costs which are 
fixed in the short run become variable 
after enough time passes, or if a large 
enough change in sales volume occurs. 
The costs of plant and equipment are 
usually fixed until the facilities wear out 
and have to be replaced. 

In dealing with such questions, the 
nature of the problem is most often cru- 
cial in determining whether a cost is fixed 



EXPANDING DEMAND 



337 



or variable. If a firm is considering how 
many trucks it should use, the purchase 
cost of its truck fleet is clearly variable. 
But when the firm wishes to know how 
best to allocate the trucks which it al- 
ready possesses, the assumption being that 
no new trucks are to be bought or old 
ones scrapped, then the nature of the 
problem dictates that the purchase cost 
of the truck fleet is fixed. 

Again, if the problem is one of long- 
range planning, certain plant and equip- 
ment costs will be variable because the 
machines will wear out during the period 
under consideration, whereas these costs 
would be fixed for a problem in, say, short- 
run production programing or control. 
In any event, the segregation of fixed 
costs requires caution, skill, and experi- 
ence; the job is an important one, but 
it is full of pitfalls for the unwary. 

Include If Separable 

The fixed costs just considered are not 
the only kind of distribution costs which 
are fixed expenses for a company. In fact, 
most fixed marketing costs are different 
in the sense that they can and should be 
allocated to sales. These are what we 
shall call the separable fixed costs. 

To illustrate: If the sales manager's 
job is not in question, his salary is, as 
we have seen, a fixed cost to the firm. 
No change in sales affects the magnitude 
of his salary (although it may affect his 
bonus). However, the time and effort 
which he spends on specific sales seg- 
ments is variable. He can spend 5 per cent 
of his time — or more, or less — working 
to promote sales in the New Orleans dis- 
trict. The more time he spends here, the 
less time he has left to spend on the 
promotion of other sales. Accordingly, we 
may be able to allocate his time with ease, 
even though his salary is a fixed cost. 



Similarly, the floor space in a ware- 
house for finished-goods inventory whose 
rental is fixed, or the capacity of a de- 
livery truck fleet, is divided among the 
different products in terms of the space 
needed by each. 

Such costs are not a direct deduction 
from the income which accrues to the 
firm as its sales increase and need not 
enter the profit calculation any more than 
do fixed costs which are common to sev- 
eral sales segments. But they are nonethe- 
less crucial for distribution cost analysis, 
for its purpose is to show where the firm's 
marketing effort can be used most effec- 
tively. Clearly, the sales manager should 
devote his time to a sales sector which 
yields $50 per hour in preference to an- 
other sector which offers only $30 return 
for an hour of his effort. 

As we shall see, it is this consideration 
which makes the methods of mathemati- 
cal programing relevant to a distribution 
cost analysis. Separable fixed costs usu- 
ally pertain to efforts or resources that are 
limited in total. 

Accordingly, the marginal cost of sepa- 
rable fixed expenses — for instance, the 
cost of the manager's time which would 
be required to make additional sales in 
each segment of the business being costed 
— should be computed if possible. And 
these figures should be kept distinct from 
variable costs. 



CHANGES IN SALES POLICY 

Suppose now it is found that, on the 
average, the sale of a dollar's worth of 
product x through medium-size retailers 
in Kansas City contributes more to profits 
than a dollar sale of product y through 
small retailers in Richmond. It is tempt- 
ing to jump to the conclusion that more 
sales effort should be allocated to the for- 
mer and less to the latter. But does this 



338 



EXPANDING DEMAND 



follow? Suppose, for example, that sales 
of product x in Kansas City have satu- 
rated the market, while the Richmond 
market is ripe for development. Clearly, 
it would not be wise to shift effort to 
Kansas from Virginia. 

Thus, the cost currently incurred by a 
specific segment of sales may be the right 
answer to the wrong question. It tells us 
how well the firm is doing now; but the 
firm wants to know whether it can do 
better in the future, and, if so, how and 
where. Accordingly, we must know the 
answer to the following hypothetical ques- 
tions: What would happen to marketing 
costs if more effort were pointed in one 
direction rather than in another? More 
specifically, how are changes in total costs 
in each sales category related to changes 
in volume in the same category? 

In effect, for each sales segment and 
each functional cost group, we want the 
kind of information contained in Figure 
4-MM. We need these data for both vari- 
able costs and separable fixed costs. Once 
these figures are obtained, the business- 
man can proceed to apportion his dis- 
tribution effort in a way virtually guaran- 
teed to increase his profits. 



TOTAL 

FUNCTIONAL 

COST 



® 




® 




® 


/(OR 
® 


VARIABLE 
SEPARABLE 
COST 

i 


FIXED) 


FIXED COST 

1 



SALES VOLUME 
Figure 4-MM 

This means that, in the simplest case 
where there is only one type of marketing 
effort in question (the more complicated 



— and more common — cases will be dis- 
cussed later), effort should be reallocated 
as much as possible to those segments of 
sales where an additional unit of market- 
ing effort will yield the highest contribu- 
tion to net profits and overhead, after de- 
duction of variable costs; that is, effort 
should be increased in that sales sector 
where there is the highest value of the 
fraction : 

additional sales — additional variable cost 

additional effort or resource devoted to 

this sector 

Further discussion and a rigorous deri- 
vation of this criterion are included in the 
appendix to this article. However, here is 
a simple illustration: Suppose the figures 
show that in Boston an additional $2 in 
the field sales force's fixed budget will 
yield an additional $10 in sales and that 
the additional variable cost of those sales 
is $4. Then the value of the fraction in 
question will be (10 — 4) -*- 2 = 3. If the 
corresponding figure for Oklahoma City 
is 2.6, it is clear that selling effort should 
be reallocated from Oklahoma to Boston. 

Often, as shown in an example in the 
appendix dealing with use of a limited 
resource (warehouse space), the proper 
allocation of effort will actually be com- 
pletely different from that indicated by 
the conventional approach via per-unit net 
profit. 



Cost-Volume Relation 

How do we find out how the functional 
costs vary with sales? The obvious answer 
is to see what has happened to costs when 
sales changes occurred in the past. But 
past events are not controlled experi- 
ments; past sales changes are often the 
result of a variety of causes, many of 
which are no longer pertinent. Moreover, 
unlike most production costs, which are 
a function of volume, changes in distribu- 



EXPANDING DEMAND 



339 



tion costs are often both a cause and an 
effect of changes in sales volume. Past 
experience may therefore be an unreliable 
test. There are, however, a number of 
techniques which will help us cope with 
this difficulty. 



MATHEMATICAL PROGRAMING 

Usually there are several types of dis- 
tribution effort or cost which a company 
must allocate carefully. The funds avail- 
able for advertising may be limited, the 
salesmen's time be fully occupied, and 
warehouse space constitute a bottleneck. 
The objective is to promote those sales 
which make the best use of all three of 
these factors. 

No one sales segment is likely to use 
all three effectively. One product may 
use advertising dollars very efficiently since 
its sales can be increased with the aid of 
relatively little additional advertising ex- 
penditure. But if this product is also 
bulky, its inventory will employ relatively 
large amounts of warehouse space. Simi- 
larly, another product may yield much 
greater profits for each additional hour of 
the sales force's time but small returns 
on each additional advertising dollar. The 
problem, therefore, is to select that com- 
bination of sales activities that will make 
optimal use of the company's facilities 
and know-how. This is what the mathe- 
matician calls a programing problem. 

What does programing do that other 
techniques cannot do? In many cases 
the standard optimization technique, that 
is, differential calculus, can, given adequate 
data, indicate precisely what is the maxi- 
mum (profit) or minimum (cost) achiev- 
able. But in some problems of optimization 
there is a complication in that the out- 
come, to be acceptable, must meet certain 
specifications which the mathematician 
calls "side conditions." 



For example, it may be most profitable 
for a firm to sell 10,000 pairs of shoes a 
week. This is, then, a sort of optimum. 
However, if the firm's warehouse can 
stock only 8,000 pairs of shoes, the op- 
timum is an unattainable goal, and it be- 
comes necessary to recompute a more 
modest and practicable target. The aim 
is to find the most profitable combination 
of outputs which do not violate the "in- 
equality side condition" that production 
must be less than or equal to 8,000 pairs 
per week. 

Programing is the mathematical method 
for analyzing and computing optimum 
output decisions which do not violate the 
limitations imposed by inequality side 
conditions. In other words, it attacks the 
same problem that distribution cost anal- 
ysis seeks to solve. Its purpose is to find 
a pattern of sales which maximizes profits 
and yet does not exceed the available 
capacities of the firm. 



Setting Up the Problem 

The mathematical programing problem 
consists of two essential parts: (1) some- 
thing which is maximized (for example, 
profit) and (2) the inequality side con- 
ditions. 

As for the first, it is generally assumed 
that the business is trying to maximize 
the sum of the gross (dollar) margins in 
the various segments of sales minus the 
sum of their variable costs. This is where 
our variable costs are used in the dis- 
tribution cost analysis. 

Next we examine the relevant inequal- 
ity side conditions. These describe the 
various limitations imposed on the firm 
by a fixed advertising budget or sales force, 
limited warehouse space, and so forth. 
There will be an inequality which corre- 
sponds to each of these limitations. For 
example, one inequality will state that 



340 



EXPANDING DEMAND 



the amount of warehouse space used by 
the various sales segments must not ex- 
ceed the available capacity. More spe- 
cifically, if, say, the warehouse space used 
for finished-goods inventories is strictly 
proportional to the sales in each category, 
we will require that the warehouse space 
used up by the different categories of sales 
add up to an amount less than or equal 
to the available warehouse space. We can 
put all of this more comprehensively and 
concisely in mathematical terms: 

ai(W) + a 2 (X) + a 3 (Y) + a^Z) < A 
bi(W) + b 2 (X) + b 3 (Y) + b 4 (Z) ^ B 
Ci(W) + c 2 (X) + c 3 (Y) + c 4 (Z) ^ C 

where A is the available warehouse space, 
B is the available sales force time, C is 
the budgeted advertising funds; and where 
ai(W) is the warehouse space used by 
W sales, a 2 (X) is the space used by X 
sales, bi(W) is the salesmen's time used 
by W sales, and so forth. (For illustration 
it is assumed that there are exactly four 
sales segments.) 

It is here our separable fixed cost fig- 
ures are used. We see now why these were 
not added to the variable cost statistics. 
The two types of figures play totally differ- 
ent roles in the analysis and appear in 
completely separate parts of the mathe- 
matical program. The variable cost repre- 
sents a deduction from the firm's income; 
the separable fixed cost represents a drain 
on its limited marketing effort; and they 
must each be treated accordingly. 

There remain two stipulations that 
must be made before the mathematician 
takes over: 

(1) We must specify that sales in any 
segment can fall to zero but can never 
be negative, or: 

W ^ 0, X » 0, Y ^ 0, and Z > 0. 

This may seem so obvious and unim- 
portant that it is not worth mentioning; 



however, it turns out that this apparently 
trivial requirement is fundamental for the 
mathematical analysis. One reason is that 
the computations are usually done by 
electronic calculators, which are not really 
as bright as sometimes supposed. They 
may do computations quickly, but they 
do not realize (unless told so specifically) 
that negative sales are out of the ques- 
tion. If some particular sales segment is 
especially unprofitable, the machine will 
figure out that the less the firm sells in 
that segment, the better off it will be. 
But, if so, why stop cutting down when 
we get to zero? Unless instructed other- 
wise, the machine is very likely to end up 
recommending negative sales in unprofit- 
able segments! 

(2) The nature of the problem may 
change so that a cost once fixed becomes 
variable and the corresponding capacity 
limitation disappears. In a planning prob- 
lem, for instance, if present warehouses 
would be inadequate for increasing sales, 
the construction of new warehouses might 
be considered by the planners, and then 
the amount of construction would vary 
with the amount of sales. The effect on 
the mathematical program is simple; the 
corresponding inequality is dropped, and 
the variable cost figures in the profit ex- 
pression are increased to include the new 
variable costs. 



LINEAR PROGRAMING 

Most frequently the word "program- 
ing" occurs in the term "linear pro- 
graming." When the facts of the situation 
state that costs will always be proportion- 
ate to sales so that, for example, a three- 
fold increase in the level of sales in any 
one segment will always exactly triple all 
the costs incurred by the segment, the 
relevant program is said to be linear. This 
is because the graph showing the relation- 



EXPANDING DEMAND 



341 



ship between sales in a sector and the 
magnitude of some cost will then be a 
straight line, as in Figure 4-MM. 

In a linear case there are great econ- 
omies in data collection and computation. 
As Figure 4-MM indicates, it takes fewer 
dots or statistics to locate a straight line 
than to plot a curved line. The few dots 
shown on the graph would obviously not 
be sufficient to permit the fitting of a 
curved line with any degree of confidence; 
yet even these few are expensive to collect. 



Limited Advantages 

Linear programing has practical uses in 
distribution cost analysis, but to appre- 
ciate them it is necessary to look first at 
its limitations. 

To begin with, it almost never fits the 
facts. If pursued to its logical conclusion, 
it is virtually certain to give the wrong 
answer. One reason is that in marketing 
various costs involve quantity discounts. 
For example, if we ship one-third of a 
carload of goods, the less-than-carload 
rates apply, and it will cost considerably 
more than one-third the amount required 
to ship a full carload. It is easy to think 
of other costs which do not rise strictly 
in proportion with sales. Inventory levels, 
and hence the costs of storage, do not 
usually rise exactly in proportion to sales 
volume; neither does the cost of sales 
management. One could make quite a 
list. 

Moreover, linear programing has an in- 
herent bias which often leads to seriously 
incorrect answers in distribution cost anal- 
ysis. It usually suggests the elimination of 
a large number of sales categories; for ex- 
ample, it may suggest that a firm end up 
selling to only a certain type of retailer 
and in only a few major cities. This hap- 
pens because of the basic theorem of 
linear programing, which, in effect, states 



that if there are, say, just three inequality 
side conditions in a problem, then maxi- 
mum profits can be obtained by confin- 
ing sales to no more than three sales 
segments. 

To illustrate, if we are trying to get the 
most out of our warehouse capacity, our 
salesmen's time, and our advertising ex- 
penditure, linear programing would lead 
us to concentrate on three sales segments 
— one which produces large profits per 
square foot of warehouse space, another 
which returns large profits per salesman 
hour, and a third which yields high re- 
turns per advertising dollar. Any types of 
sales which are second best to these three 
would be eliminated. 

Of course, this is all wrong. For one 
thing, there are diminishing returns to 
effort; markets become saturated. If one- 
twentieth of the national advertising 
budget is used in Chicago, moving the 
entire advertising effort to Chicago will 
not increase sales there twentyfold. The 
transfer of effort would be a good thing 
only up to a point, and it would be most 
profitable to devote much of the adver- 
tising budget to other market areas. 



Best Direction 

All of this is not to say, however, that 
linear programing has no uses. Quite the 
contrary. While a linear program usually 
will not compute a correct optimum be- 
cause the changes it suggests will go too 
far, there is yet a very strong presumption 
that it will correctly indicate the best 
directions of change. It will identify cor- 
rectly the sales segments to which more 
effort should be allocated. Accordingly, 
where it is too expensive to undertake a 
full-scale nonlinear programing analysis, 
or where the necessary data are simply 
not available, linear programing can still 
be exceedingly helpful. 



342 



EXPANDING DEMAND 



CONCLUSION 

In practice, the great bulk of the sav- 
ings which distribution cost analysis has 
made possible can be ascribed to its abil- 
ity to find those glaringly misallocated. 
Using this technique, many sales man- 
agements have been able to redirect their 
efforts and achieve very substantial addi- 
tions to profits. 

As most sales executives know, the data 
available are rarely so accurate that the 
analysts can trust a very refined calcula- 
tion. But when their information points 
out a very costly case of misdirected effort, 
they can confidently proceed to take 
remedial action, for no small error in data 
and computation will normally account 
for the considerable losses which distribu- 
tion cost analysis often reveals. In sum, 
even fairly inaccurate data and fairly crude 
computational techniques will usually turn 
up those inefficiencies that are most glar- 
ing and constitute the most important 
opportunities to increase profits. 

Accordingly, even though they may not 
have the data, the time, or the budget for 
a mathematical programing analysis, man- 
agers usually can still obtain much of the 
benefit of a full-scale distribution cost 
analysis by using a linear programing ap- 
proximation or just the general approach 
outlined in this article. They can increase 
the accuracy of their information with 
only inexpensive modifications in distribu- 
tion cost analysis procedures, as we in- 
dicate in the appendix which follows. 
Thus, the principles alone of mathemati- 
cal programing will serve, when used in 
conjunction with a distribution cost anal- 
ysis, to help management achieve from its 
marketing operations a marked increase in 
efficiency and substantially higher profits. 



To what extent can the techniques de- 
scribed here be used without accurate 
data on cost-volume relationships? Will 
they help more with problems of how 
much to spend in marketing effort, or 
with problems of where and how to 
spend it? 

APPENDIX — SOME SUGGESTIONS FOR 
COMPUTATIONS 

A. Profit computation for allocation 
decisions 

When the firm is considering the allo- 
cation of the currently available amount 
of some type of marketing effort, it is 
posing a question which requires that 
the expenditure on that type of effort be 
treated as a fixed cost. In posing such a 
question the analyst is told, in effect, that 
the firm has, let us say, so many sales- 
men at its disposal, and he is then asked 
how the members of the sales force can 
best divide their time among products, 
customers, territories, and so forth. 

In distribution cost analysis, the vari- 
ous separable fixed costs (say, the ware- 
house floor space) allocated to each sales 
segment are translated into money terms, 
and the sum of these and the allocated 
variable costs are subtracted from the dol- 
lar gross margins to obtain a net profit or 
loss figure for each segment of the busi- 
ness. The implication is that marketing 
effort ought to be decreased in the sectors 
which show a large net loss and increased 
in the sectors having the highest per-unit 
net profit figures. In many cases, however, 
this implication may be incorrect and 
misleading. 

For simplicity, we will illustrate this 
with the linear case where variable and 
separable fixed costs per unit for any sales 
sector do not change with the volume of 
sales in that sector. We will also assume 
that warehouse space is the only separable 



EXPANDING DEMAND 



343 



fixed cost. We will deal with the sales W 
and X in two sectors. 

Now, it is usually suggested that effort 
should be transferred from X to W if 
and only if W's per-unit net profit is 
greater than X's; that is, if and only 
if (Formula I) : 

Gross margin of W — variable cost of 
W — separable fixed cost of W > gross 
margin of X — variable cost of X — 
fixed separable cost of X. 

But Formula I does not always cor- 
rectly indicate the most profitable real- 
location of marketing effort. The most 
profitable thing to do is to increase the 
effort devoted to the sector which yields 
the highest contribution to profit and 
overhead per dollar of warehouse space. 
That is because warehouse space is the 
separable fixed cost — the effort or re- 
source factor which needs to be econo- 
mized — and the best profit-making strategy 
is to use that space where it earns the 
most. This proposition is easily proved: 

If A is total available warehouse space and 
is completely utilized, we have: 

A = A W W + A X X 

or: X = A/ A x - A W W/A X 

where A w and A x are the respective number 
of square feet of warehouse space used up 
per unit of W and X. Now let T w and T x 
represent the contribution to profit plus over- 
head per unit of W and X respectively. Sub- 
stituting the expression just obtained for X, 
the total contribution to profit plus overhead 
of both sectors then will be: 

WT W + XT X = WT W + A/A X T X - A W W/A X T X 
= A/A X T X + W(T W - A W /A X T X ). 

Thus an increase in W will add to this 
figure if and only if T w — A W A X T X > o; i.e., 
if T w /A > T x /A x . To put it in another 
way, an increase in W will add to the profit 
figure only if T w /A w , the contribution per 
unit of warehouse space from stocking W, is 
greater than T x /A x , the contribution per 
unit of warehouse space from stocking X. 



The significance of all this for our 
main question is that actually more effort 
should be allocated to W from X if and 
only if the figure for W's contribution to 
profit and overhead in ratio to W's use 
of warehouse space is greater than the 
corresponding figure for X; that is, if and 
only if (Formula II): 

(Gross margin of W — variable cost of 
W) -5- separable fixed cost of W > 
(gross margin of X — variable cost of X) 
-*- separable fixed cost of X. 

Now the important thing to note is 
that Formulas I and II will not always 
point in the same direction. For example, 
if both gross margins are equal to 1.0, 
if the unit variable cost of W is 0.3 and 
the separable unit fixed cost of W is 0.2, 
and if the variable cost of X is 0.1 and 
the separable fixed cost of X is 0.3, then 
the net profit computation under Formula 
I yields less unit net profit for W, since: 

1.0 - 0.3 - 0.2 = 0.5 < 1.0 - 0.1 - 0.3 = 0.6. 

Thus, distribution cost analysis, as often 
practiced, would infer that effort should 
not be shifted from X to W, or that the 
shift should be in the other direction. 
But the correct computation under For- 
mula II shows that W's contribution to 
profit and overhead per unit of warehouse 
space is really greater than X's, because: 

(1.0 - 0.3) tO. 2 = 3.5> (1.0 - 0.1) ^0.3 = 3. 

In other words, the firm should shift ef- 
fort from X to W, in precisely the oppo- 
site direction from that indicated by the 
usual net profit computation of For- 
mula I. 

It should be remembered, however, that 
the procedure which has just been shown 
to be correct is valid if there is only one 
category of fixed cost and one corre- 
sponding facility to be allocated optimally. 



344 



EXPANDING DEMAND 



Where the number of facility limitations 
is greater than this, it is necessary to 
utilize the standard mathematical pro- 
graming computations discussed earlier in 
this article. 



B. Marginal versus average data 

As we have seen, average costs per 
unit cannot tell us how to reallocate effort. 
The information needed to decide 
whether a shift of effort will be profitable 
is the addition to costs and sales which 
will result from such a shift; that is, the 
marginal costs and revenues. In practice, 
however, it is much easier and less ex- 
pensive to obtain average gross profit and 
cost figures than to acquire the corre- 
sponding marginal figures. Most distribu- 
tion cost analysis, therefore, employs aver- 
age data in its calculations. We believe 
that this often yields satisfactory approxi- 
mations, but good practice requires that 
the data be corrected for the most glaring 
divergences between the marginal and 
average cost figures. In many cases it should 
be possible to recognize such differences 
fairly easily. 

Here economic analysis provides us 
with several simple rules: 

(1) All fixed costs must be eliminated 
from the average cost figures. In partic- 
ular, if there are setup costs which must 
be incurred in order for the firm to op- 
erate at all, these must be omitted. A 
perusal of the accounting practices of the 
firm may indicate gross profit or cost fig- 
ures in which such costs enter heavily, 
and appropriate corrections should be 
made. 

(2) If demand does not respond read- 
ily to effort, marginal revenues will be 
considerably lower than average revenues. 
In effect, this is because there will be 
diminishing returns to effort in such mar- 
kets, so that additional effort will nor- 
mally yield smaller revenues than does 



most of the present effort. This means 
that if experience suggests that market M 
is pretty much saturated and will not re- 
spond readily to further marketing effort, 
the average gross profit figure for this 
market must be reduced to obtain a num- 
ber closer to the marginal gross profit. 

In a few cases, marginal revenues will 
be greater than average revenues because 
additional selling activity can increase the 
effectiveness of effort already expended. 
In such cases an upward adjustment of 
the average gross profit is called for. 

It will not be possible to identify all 
of the sales segments requiring this sort 
of adjustment or to determine the precise 
magnitude of the required changes. But 
with a little experience one usually can 
recognize, on the basis of interviews with 
management and examination of the 
available records, the most extreme cases 
of divergence between marginal and aver- 
age gross profit; the analyst can then 
eliminate the most serious sources of 
error. 

(3) On the cost side we have a similar 
rule. Where there are seriously diminish- 
ing returns, increase the average costs; if 
demand does by any chance respond more 
readily to added sales effort, reduce the 
average costs. This rule applies in the 
same way to production and to distribu- 
tion costs. 

(4) The magnitude of the required 
change in an average gross profit figure 
and in an average cost figure can be in- 
ferred with the help of the following 
standard formulas: 

(a) Marginal gross profit — average gross 
profit -f- sales X rate of change in gross 
profit per sale. 

(b) Marginal cost = average cost + sales 
X rate of change in average cost per 
sale. The rate of change in average cost 
per sale would mean, for example, 
(average cost in sector X when sales are 
$1,001,000 — average cost when sales 
are $1,000,000) -s- the increase in sales 
[$1,000]. 



EXPANDING DEMAND 345 



BIBLIOGRAPHY TO SECTION FOUR 

Asplev, John, editor, The Sales Manager s Handbook. Chicago: Dartnell Corpora- 
tion, 1958. 

Beckman, Theodore N., Nathanael H. Engle, and Robert D. Buzzell, Whole- 
saling. New York: The Ronald Press Company, 1959. 

Clewitt, Richard M., editor, Marketing Channels for Manufactured Products. 
Homewood, 111.: Richard E. Irwin, Inc., 1954. 

Duncan, Delbert J., and Charles F. Phillips, Retailing: Principles and Methods. 
Homewood, 111.: Richard D. Irwin, Inc., 1959. 

Katz, Elihu, and Paul F. Lazarsfeld, Personal Influence: The Part Played by Peo- 
ple in the Flow of Mass Communications. New York: The Free Press of 
Glencoe, 1957. 

Klapper, Joseph T., The Effects of Mass Communications (survey of communi- 
cations research). New York: The Free Press of Glencoe, 1960. 

Maslow, Abraham H., Motivation and Personality. New York: Harper & Brothers, 
1954. 

Russell, Frederic A. and Frank H. Beach, Textbook of Salesmanship. New York: 
McGraw-Hill Book Company, Inc., 1955. 

Smith, George H., Motivation Research in Advertising and Marketing. New York: 
McGraw-Hill Book Company, Inc., 1954. 

Snygg, Donald and Arthur W. Combs, Individual Behavior. New York: Harper 
& Brothers, 1949. 



SECTION FIVE 



Competitive 
Position 



In this section we shall concentrate on the way a company seeks to gain a larger 
share of the market and on that aspect of marketing research which is most crucial 
in connection with brand preference and other indications of competitive position. 



PRIMARY VS. SELECTIVE DEMAND 

There are two ways for a company to gain more business: (1) to stimulate 
primary demand for whatever it has to offer, so people actually buy more of this 
kind of product in general, and (2) to stimulate selective demand for this partic- 
ular brand. Thus, the California Orange Growers Association, cited in Section 
Four, first put its efforts into building demand for oranges— all oranges; but sub- 
sequently switched to promoting the Sunkist brand as better than other brands 
("more juice," and so forth). The same distinction can apply to stores. The 
Supermarket Institute publicizes the role of supermarkets— all supermarkets; but 
individual chains or stores advertise, give trading stamps, and so on, to get shop- 
pers to come to their particular supermarkets. 

In a very real sense, however, the distinction between primary and selective 
demand is arbitrary. The amount of resources (whether money or anticipated 
money, that is, credit) that people have at their command is limited, so demand 
for anything, even the broadest classes of products and services, is relative; it exists 
in comparison with demand for all other things; and, if it builds up faster than a 
person's total resources, that person must be abating his demand for something 
else. This is true even though, in the long run, demand, by its cumulative effect 
on the economy, can enlarge many individuals' resources. 

Think of the different gradations of demand, from the most primary to the 

347 



348 COMPETITIVE POSITION 

most selective, whereby people (living comfortably above the level of physical 
need) choose between alternatives: 

First of all, between spending or saving. 

If spending, then between food or clothes. 

If food, between green vegetables or potatoes. 

If green vegetables, between peas or string beans. 

If peas, between frozen peas or canned peas. 

If canned peas, between Del Monte brand or Libby brand. 
(The gradation could also be in terms of stores— to shop or not to shop, to go to 
a food store or a clothing store, to go to a chain store or a neighborhood grocer, 
to go to an A & P store or a Safeway store, and so forth.) 

In other words, at every stage, selling depends on the buyer's being persuaded 
that taking a particular action is better than not taking that action. So, in a sense, 
all selling is selective or competitive; and all buying is, too. Even a so-called 
monopoly like the American Telephone and Telegraph Company is competing 
against other uses for people's money. An extension phone competes against some 
other kind of convenient appliance, and a long-distance call competes against 
telegraph or special delivery mail— or against saving the money toward something 
else like a trip or a gift. 

At the lower end of the economic scale, or in countries that are at an earlier 
stage of development, the choice is between products or services that are sharply 
different in what they offer (for example, shoes vs. medicine). But, to an increas- 
ing extent today, both here and abroad, the choices are more likely to be between 
products that offer somewhat similar satisfactions (for example, pie vs. cake, or 
TV vs. moving pictures) or between brands of very similar products (for example, 
Pepsi-Cola vs. Coca-Cola). 

And the finer these distinctions between products or brands become, the 
more the buying turns on subtle points of predisposition and persuasion, which 
are smaller in effect, and less predictable and measurable. So marketers need to be 
able to appraise where they stand in reference to competitors at any given moment 
and in reference to where they stood before (that is, how much progress they are 
making). 

SHARE OF MARKET 

Company sales figures are not sufficient for measuring progress or competitive 
effectiveness. Because of inventory build-ups and rundowns at the wholesale and 
retail levels, current company sales are not the same as current sales to consumers. 
Even current sales to consumers are not an indicator of trends that still are in the 
making, in terms of the current effect of advertising and other marketing moves 
which often does not reveal itself immediately in buying action. Even more im- 
portant, unless sales are compared with competitors' sales or against total industry 



COMPETITIVE POSITION 349 

sales, the company has no way of knowing whether it is doing as well as it could 
do or ought to do in terms of the share of the market it is securing and the poten- 
tial that is still there to be won. 



Taking Measurements 

There are a number of ways in which marketers can improve their knowledge 
in this situation. First of all, there are various services available which measure 
the actual movement of goods at the consumer level. Two of the best known are: 

1. A. C. Nielsen Company, on the basis of actual store audits made by 
trained personnel, with the cooperation of store managements in allowing 
examination of inventories, invoices, and so on. Clients get a clear picture of 
their sales vis-a-vis competitors. (See Case 5-4.) 

2. Market Research Corporation of America, on the basis of a nation- 
wide panel of over 5,000 consumers, who keep diaries of their family pur- 
chases. Careful procedures are followed so panel members will make accurate 
reports, but there is likely to be somewhat less precision than with store 
audits. On the other hand, the diary information affords more kinds of quali- 
tative information, such as what kinds of families are buying and their pattern 
of purchases (new buyers, former buyers, repeat buyers, frequent buyers, and 
so on). Figure 5-A is a chart prepared by MRCA to illustrate its services. 
There are also a number of special services available, such as drugstore audits 

of doctors' prescriptions and audits of building permits as a measure of construc- 
tion activity. Likewise, some large companies or companies with problems not 
covered by outside services, and also some advertising agencies and publications, 
maintain their own special consumer panels or audit services. 

For advertising, in particular, there are services which measure on a sampling 
basis and report the number of people reading, hearing, or viewing advertising 
messages. Thus: 

Starch ratings— -Daniel Starch and Staff have for many years (since 1932) 

been obtaining three measures of readership of space advertising (both for the 

whole ad and for the various parts of it like headlines, illustration, text, and 

company signature) : 

1. NOTED— The percentage of all readers of the current issues interviewed 
who remember that they previously have seen the advertisement in the 
particular publication. 

2. SEEN-ASSOCIATED-The percentage who remember that they have 
seen the advertisement in the particular issue concerned and associate it 
with the name of the product or advertiser. 

3. READ MOST— The percentage who have read 50 per cent or more of the 
text of the advertisement. 

In every Starch report it is pointed out that the percentages contain an 







<u co 



• c § 

-t-> 0) 4h 

C Jh *? 

Q) *H 9 

^ 3 g 

h o o 

u o a 



- 3 

+3 O 3 

g 0> « CO 

Pi > C S 



< 




TJ 


en 


C 


S-. 


as 


cu 


pq 


a 

R 


3 


w 

a 





( > 


H 





Cj t*H ? CO 

3 S c 



a 



o 

^ CO 

O"0 it 

° .2 2 

Pi 73 6 

_ <u o 

S&5 






o 5 <j 

co > = 

cu < -a 

TO crt 



U O 



(2 U 



COMPETITIVE POSITION 351 

element of error (they are based on small samples); that averages for several 
ads in a given campaign are more stable than figures for individual advertise- 
ments; that the purpose is to aid in evaluating the types of copy, headline, 
illustration, use of color, page size, and so forth, which do the best job of 
attracting and holding readers; and that the report is not a basis for making 
comparisons between publications. Used in this light, they can be very helpful. 

Nielsen (originally Hooper) ratings— These are based on a sampling of 
telephone homes; respondents are called and asked if anyone in the home is 
listening to the radio or watching TV; and, if the answer is yes, he is asked 
what station and program, and what sponsor or product. Calls are made co* 
incidentally— that is, to many homes at the same time— so ratings will express 
the per cent of radio or TV homes tuned to various programs at any given 
moment, and so, presumably, the comparative position of the advertising 
carried by the programs. 

According to Harper W. Boyd, Jr., and Ralph Westfall, this system has 
the advantage of speed and economy, but people in rural homes and people 
listening and watching after 10:30 p.m. are neglected; the effect of tuning in 
and out is not taken into account; and these procedures do not produce any 
continuous information about the audience. One cannot tell how many 
homes are reached over a period of several programs— that is, what the cumu- 
lative audience is. 1 Again, these limitations do not mean that such tech- 
niques are not useful (advertisers would be lost without them), but only 
that they must be used carefully. 

There are also a number of other radio and TV services like Trendex, the 
Pulse recall method, the Nielsen Audimeter (an electronic device which keeps 
a complete record of what a specific set is tuned to), and Arbitron (with instan- 
taneous computing of channels being watched, in major cities). 

The basic difficulty with all such methods, above and beyond any inade- 
quacies in the procedures used, is that they measure only superficial and partial 
indications of how effective either the media carrying the advertising or the ad- 
vertising messages themselves are. All they measure, really, is how many people 
have been exposed, in varying degrees, to the possible effect of the advertising. 
The degree to which the exposure produces action (or a change of attitude in the 
direction of action) depends on who is exposed, why they are receptive or not, 
and how they are affected. As pointed out in Section Four, people's exposure to ads 
will produce different results depending on their predispositions. Also, there will 
be differences both between kinds of programs and between kinds of advertising 
objectives. Thus, exposure to a symphony program on TV and exposure to a 
Western on the same channel are not similar phenomena even if watched by 
identical people, and the products sold by the companies sponsoring them do 
not call for the same kind of advertising effects. If marketing action is to be taken 



1 Marketing Research (Homewood, Illinois, Richard D. Irwin, Inc., 1956), p. 561, 



352 



COMPETITIVE POSITION 



on the basis of measures of exposure (or any other kind of measures), then, as 
always in marketing, analysis must be applied to the particular product-market 
situation. 



Changes in Attitude 

What we need to know is what is happening inside people's minds— what 
their attitudes are toward the product, or what their particular reaction is to the 
advertisement. In general, if we look at a cross section of the various stages through 
which people move as they progress toward buying action, we will find that each 
closer measure of favorable attitude is smaller in number of people but surer as 
an indicator of final action; and that for similar product-market situations, the 
changes are proportionate. Thus, for two products, A and B, at a given time 
the picture would tend to be: 



Ad readership or viewing 



W% 



60% 



Ability to recall or 
recognize brand 



6i% 



40% 



Preference for brand 



A 



45% 



25% 



Intention to buy 



30% 



15% 



Actual buying 



20% 



10% 



Of course, the drop-off will vary for different product-market situations, and 
there is many a slip between the ad and the cash register, so this tendency toward 
proportional progression is not always helpful in predicting the absolute rate 
of sales from one of the intermediate measures. However, used with caution, it 
can be helpful in comparing the relative effectiveness of the promotion for several 
products (either within the company's own line or between the company and its 
competitors). 



COMPETITIVE POSITION 353 



EVALUATION OF ADS 

As for what the value of a specific ad is in getting through to people and 
bringing about a change in their attitudes, there are a number of possible re- 
search procedures. The degree to which people recall an ad probably does reflect 
their interest in it and its impact on them. On the other hand, there are other 
indications of interest and effectiveness, ranging from simple awareness of the 
brand of the product to outright desire to purchase that brand rather than another. 
Simply asking people what they think of an ad is hardly enough; hence, many 
methods of comparing advertising effectiveness involve exposing people to a series 
of ads under realistic conditions of limited time with the order rotated from inter- 
view to interview to avoid the influence of order, then asking them what they 
remember, and so on. 

Richard D. Crisp has been a strong proponent of the playback method. As 
he describes it, a playback is simply a verbatim report, in the respondent's own 
words, of everything he or she can recall about an advertisement. It is illustrated 
by an advertisement for Kix cereal. The ad showed two pictures of stenographers 
at their desks. One was bright-eyed and chipper, obviously alert and happy; the 
other had her head on her hands, and looked dopey and groggy. Readers were in- 
vited to guess which girl had the breakfast including Kix. The girls were named 
"Lupe" and "Droopy." Here are four actual playbacks illustrating the range of 
content from the same ad: 

1. It was a Kix ad. I just looked at it to see if I was interested or not. 

2. Kim or Kix cereal. There were two pictures of the same lady, but I 
don't remember what she was doing or what the ad said. 

3. Kix cereal. Difference in woman who had it, the pep and energy it 
gave her, and how sluggish the other one was that didn't have it. Package 
of Kix with a dish on it showing what it looked like. 

4. There was an ad with two girls on it. It said which one is Loop and 
which one is Droop. It was for a breakfast food, but I don't remember what 
brand. I just skimmed the ad. I noticed one girl was alert, and one was look- 
ing the other way. 

This range of playbacks illustrates the analytical problem. If all respondents 
reacted like the one in playback No. 1, you'd junk the ad. If they were all like 
No. 3, you'd be delighted. This illustrates the quantitative and qualitative prob- 
lem involved in analyzing playbacks. It is necessary in evaluating an advertise- 
ment first to set up meaningful frameworks for groups of playbacks; and then to 
determine quantitatively how many playbacks fall into each group. In compari- 
sons of ads, both the quantitative and qualitative aspects of the playbacks would 
guide your appraisal. 2 



2 Richard D. Crisp, Marketing Research. (New York: McGraw-Hill Book Company, 1957), 
pp. 630-631. 



354 COMPETITIVE POSITION 

Suppose three ads are being compared, and five qualities of response (to 
playbacks or other methods of securing reaction) can be roughly identified, as 
follows: 

Ad A Ad B AdC 



Very favorable 


35% 


30% 


5% 


Moderately favorable 


5 


40 


70 


Neutral 





15 


10 


Moderately unfavorable 


20 


10 


10 


Very unfavorable 


40 


5 


5 




100% 


100% 


100% 



Which ad is likely to be the most effective? To answer that question depends 
on some further quantitative estimate of what the qualitative ratings mean in terms 
of probable sales, such as: 

Very favorable = 30% probability of purchase 

Moderately favorable = 15% 

Neutral = 10% 

Moderately unfavorable = 5% 

Very unfavorable = 0% 

If the quality-of-response percentages are multiplied by the probability-of- 
purchase percentages, and the resulting figures added, Ad B turns out to be clearly 
the best with a total score of 17.00, while C is a poor second with 13.50, and A 
trails with 12.25. It should be noted, however, that the probability percentages 
require some very astute estimating. It does not necessarily follow that there is a 
nice even gradation from favorable to unfavorable; it is even conceivable that a 
strong unfavorable reaction may be less damaging to sales prospects, just because 
it indicates positive interest, than a neutral reaction, which indicates apathy. In 
other words, management must think through (or, if practical, research) the 
particular product-market situation in order to arrive at the probability figures 
to use. 

Even better than basing decisions on reactions to ads, no matter how thor- 
oughly and wisely done, is the sales-testing of ads under controlled conditions. It 
is not always practical to try out ads in this way, because of the difficulty of mak- 
ing the proper arrangements when the media involved are not local in character; 
use of handbills or posters is not an exact substitute for an ad in a magazine, for 
instance. And the difficulties are compounded when the ad is part of a campaign for 
a new product, since then special efforts must be made to get the product stocked, 
and there is the disadvantage of alerting the competition sooner than would other- 
wise be necessary. 



COMPETITIVE POSITION 355 



Economic Value of Research 

But the cost, in terms of money, time, and effort, may still be worthwhile if 
the amount of money at stake in the product is large. There is much to be said 
for management's attempting to be explicit in figuring out what the stake is in the 
case of a proposed research product. Here, again, probabilities afford a useful 
approach. Take these estimates: 

Present value of operating profits on expected sales over the prod- 
uct's life (handled as if product development and marketing 
costs of introducing the product are a capital investment) . . . .$1,000,000 

Role of marketing in establishing product on the market 60% * 

Advertising campaign as part of total marketing effort 20% 

This particular ad's part in advertising campaign 15% 

Dollar value of ad (.15 X .20 X .60 X $1,000,000) $ 18,000 

Now suppose management makes the probability estimates shown in Table 
5-B. The possible outcomes are the degrees to which the ad may be successful in 
performing its potential $18,000 role, depending on the research. The probabilities 
express the chances that the three alternative research courses— (1) a $4,000 sales- 
testing project, (2) a $1,000 experimental game project, and (3) no project at all, 
or simple use of management judgment— will select an ad of that degree of suc- 
cessfulness. It turns out that in the assumed conditions the $4,000 market research 
project is not as likely to pay off as use of management judgment, but the less 
ambitious $1,000 project is the best bet of all. As suggested earlier, this might be 
the kind of game, simulating real buying, where a sample of people are given a 
certain amount of token money; are shown an array of ads; and then, under rules 
designed to make their actions as close to what they would be under actual market 
conditions as possible, spend the token money for their choice of the products 
advertised. 



♦Perhaps estimated like this: 



Product development 




25% 




Delivery and service 




15% 




Marketing 






Advertising 


Getting company name known 


5% 




2% 


Identifying prospects 


5% 




1% 


Arousing interest 


10% 




2% 


Introducing the salesman 


10% 




2% 


Building confidence 


10% 




5% 


Making final sale 


20% 




- 






60% 


12% 1 



12% ^60% = 20% 
Total 100% 



356 



COMPETITIVE POSITION 



TABLE 5-B. DETERMINATION OF OPTIMUM RESEARCH PLAN ON AN AD 
WITH ECONOMIC VALUE OF $18,000 



Probabilities 



$4,000 
Research 

.90 
.10 



$1,000 
Research 

.25 
.70 
.05 



Possible 
Outcomes 

100% 

90% 

80% 

70% 

60% 
Total 
Cost 
Net Probable Payoff 



♦Computed: Outcome x Probability x $18,000. 



Management 
Judgment 



.05 
.15 
.60 
.15 
.05 



$4,000 
Research 

$16,200 
1,620 



$17,820 

-4,000 

$13,820 



Dollar Equivalents' 



$1,000 
Research 

$ 4,500 

11,340 

720 



$16,560 

-1,000 

$15,560 



Management 
Judgment 

$ 900 

2,440 

8,640 

1,890 

540 

$14,410 

- 

$14,410 



Actually it would be sounder to make each of the estimates used above a 
range rather than a fixed percentage, say, 50 per cent to 70 per cent instead of 
60 per cent. The final figures would then also be ranges, of course. This would be 
more realistic, and would also serve to call sharper attention to the fact that the 
final figures were the results of estimates. 

There is no intention here to suggest that a decision to do or not to do re- 
search should be made only on such a mathematical basis as above, or that it is 
worthwhile to subject all research proposals to this kind of calculation. However, 
it may be helpful to think through the question of the economic value of mar- 
keting research from time to time, if only to get and keep some perspective on 
whether the company's marketing research effort as a whole is roughly in line with 
the contribution it makes. Also, if what is being considered is a part of a larger 
continuing project, the value may be greater because of its repetitive use (either 
"as is" or with some judgmental modification). 

Of course, another value of the experimental approach— that is, testing ef- 
fectiveness by actual sales— is that it would refine the accuracy of some of the es- 
timates in a probability calculation like the one just described; and it might be 
that, with accumulation of experience, certain categories of product-market situa- 
tions could be developed and used as a guide for estimating the economic value of 
research (but always with a built-in question about the possible existence of dif- 
ferences in the particular situation). 



THE COMMUNICATION ELEMENT 



The above observations apply also to research in connection with other mar- 
keting moves than advertising. But advertising, especially, brings up the need to 
plumb the attitudes and opinions of people, because it works almost wholly 



COMPETITIVE POSITION 357 

through causing change within people. Advertising, selling, and promotion in 
general provide the marketing moves that serve to actuate the communication 
element of the marketing mix; they are designed to take the appeals provided by 
the other elements— product characteristics, physical location, and price— and to 
bring them into people's minds where, consciously or unconsciously, they move 
people's attitudes closer to buying. 

It is true that, to some, and often a considerable, degree, the other elements 
have their own built-in potentials and limitations. Thus: 

Refrigerators with extra-large freezing units appeal to those who have 
reason to want that particular product characteristic (say, large families, with 
access to fresh produce), and can hardly be made to appeal to those who do 
not. 

Low-price meats appeal to those who must watch their pennies; high-price 
meats cannot be bought by those who can't afford them— bargain-hunters and 
spendthrifts being the exceptions that always keep marketers from daring to 
take anything for granted except exception itself. 

Grinding wheels in industrial supply firms that are located within the 
distance of a short truck trip and carry a wide range of types and sizes will 
appeal to machine shops, and will not appeal if such convenience of service 
does not exist because of the manufacturer's distribution policy. 

It is also true that these same kinds of appeals can serve, not just for primary 
expansion of demand, but for the most selective kind of promotion. One refriger- 
ator company will take business away from its competitors by having a larger 
freezing unit; one food store will build patronage by offering lower-price meats; 
one mill-supply firm will attract new customers and win a larger proportion of the 
business of its existing customers by stocking a wider variety of sizes and kinds of 
grinding wheels and providing prompter delivery— or a manufacturer of grinding 
wheels will increase his share of the market by getting his product stocked by the 
industrial supply firms that provide such better service. 

Even in such cases promotion also plays a part because it communicates these 
appeals to people: it describes the characteristics, announces the price, calls at- 
tention to the distributors. But in selective expansion of demand, where the dif- 
ferences between competing brands are small just because of the force of compe- 
tition, promotion has to do more than inform. 

Giving More Meaning 

It has to take some feature of a product that represents an advantage, how- 
ever minor, however unsubstantial, however unrecognized, and make it seem 
larger; that is, give it more meaning in people's minds. It seeks to make people want 
that particular brand no matter what product characteristics competing brands 
have, no matter what the relative prices, no matter where other brands are availa- 



358 



COMPETITIVE POSITION 



ble. A store has the same problem: to make some feature or features of its service 
stand out in people's minds so they come to it rather than to other stores that, 
after all, carry much the same merchandise at much the same prices. 

Let us look a little more closely at the way promotional communication works 
in conjunction with the other elements: 

IN CONJUNCTION WITH PRODUCT CHARACTERISTICS. Man- 
ufacturers or storekeepers must continually add new features. However, the 
immediate advantage does not last long. Competitors are never far behind; 
and the more successful a new feature, the quicker it will be followed by com- 
petitors (see Figure 5-C). But there is a cumulative favorable effect for the 




2 3 4 5 6 7 8 

YEARS AFTER FIRST ADVERTISEMENT OF FEATURE 



Figure 5-C 



company that innovates regularly; and, above all, the new features provide 
material for promotion to use in building a preference for the brand. This 
opportunity to advertise actually may sometimes be more important than 
the inherent attractiveness of the new features themselves. 

The race to have something new and different with which to attract 
customers away from competitors is responsible for the often criticized prac- 






COMPETITIVE POSITION 359 

tice of getting people to feel that they must buy new automobiles and other 
goods long before their present ones are worn out— "planned obsolescence," 
it is often called. The fact remains that the competitive drive to keep ahead 
is what does succeed in keeping designers alert, and thus, over time, causing 
substantially improved products to be delivered to the public. Further, the 
fact that the marketers engage in constant competitive advertising raises 
the level of consumption of all brands in an industry and thus, adding all in- 
dustries together, total consumption, total employment, total wages, and total 
purchasing power. 

IN CONJUNCTION WITH PRICING. The pricing area offers an- 
other example of the way in which promotion often carries the main burden 
of competitiveness. When companies are striving for the patronage of con- 
sumers with products that are not very different, naturally prices tend to be 
the same. Why should anyone pay substantially more for a Ford than for a 
Chevrolet, for Maxwell House coffee than for Chase and Sanborn? When 
one company lowers the price, its competitors are almost bound to follow— 
unless the one company has some marked cost advantage, which seldom is 
the case. 

Not that prices aren't lowered from time to time. There is a temporary 
advantage in having a lower price; and, just as with product features, if a 
company gets a few new people to try its product every time it can be bought 
for less, then, in the long run, there is a favorable cumulative effect. But, 
again, as with product features, the "special" price often serves simply as 
ammunition for advertising and other promotion; and this opportunity may 
actually be more important than the price reduction itself. 

IN CONJUNCTION WITH PHYSICAL LOCATION. In a sense, 
promotion has the same over-riding role in connection with the physical 
location where products are offered. Much effort can and must be exerted to 
get the product stocked (and displayed) in the right kind and number of 
outlets, but often this is just to set the stage so advertising can go to work. 
When the manufacturers of Camel cigarettes were advertising "I'd walk a 
mile for a Camel," they also were seeing to it that people could buy Camels 
cmywhere that any brands were sold, so the influencing of smokers of other 
brands to switch to Camels would be that much more likely to be effective. 

It should be noted that different marketers will still use different combinations 
of the elements of the marketing mix, including the communication element. 
Water heaters and wrist watches will use but few new features, in contrast to 
automobiles and cameras, for example, simply because they require little skill and 
attention to operate, and therefore, in their case, new features have less interest, 
less meaning, less selective effect. 3 Large products purchased infrequently, like sets 



3 See John B. Stewart, "Functional Features in Product Strategy/' Harvard Business Review, 
March-April 1959, p. 67. 



360 COMPETITIVE POSITION 

of sterling silverware, will hardly gain in competitive attractiveness by a price re- 
duction of a small amount, and may lose because of the cheapening effect. One 
brand of sugar will not sell much better than competing brands if pushed by sales- 
people in a few selected outlets, but it will respond to a price cut; while the oppo- 
site would be the case with a make of oil burner. 

The point is, however, that in all cases, the more selective the situation (that 
is, the more buyers must choose among closely similar products and services), the 
more competitive success depends on the skill with which the communication 
element is exercised— in short, on the quality of the promotional effort itself, 
rather than the appeals inherent in the product or its physical conditions. 

And here, too, there are both potentials and limitations. As far as the limita- 
tions on communicability of appeals are concerned, we have already seen, in con- 
nection with the earlier discussion of how much to spend on advertising, that one 
product will be less promotable than another product if the assurance of quality is 
less important to people because of the way they buy it or use it. Even more 
significant at this point is the fact that certain kinds of people are less responsive 
to communication. But this also provides an especially useful tool for selective 
expansion of demand, when the coin is turned over and we go ahead to consider 
the potentials. For one marketer can gain advantage for his product over his com- 
petitor's by improving the communication process itself. 

There are two ways to improve the communication process: 

(1) Through the skill of specialized practitioners in words and art and 
other creative presentation, such as can be found in advertising agencies. This 
is not within the purview of this book. 

(2) Through the strategy of focusing the appeals more sharply and 
surely, which is something that marketing management should be concerned 
with. 

Here we are considering the communication of appeals as a mechanical proc- 
ess, like the transmission of light. Light can be refracted into many small rays or 
focused into a few strong beams. The brighter the light which is the source, the 
greater the area you can cover; the weaker the light, the more you need to con- 
centrate it. Also the more photo-sensitive the material on which the light falls, the 
less clear its transmission has to be to secure a given effect. 

In this analogy, the brightness of light is the strength of the appeals that the 
product has; the sensitivity of the material on which the light falls is the predispo- 
sition of the people to receive the appeals, feel them, like them, and react to them. 

Behavioral science supplies theory for conceptualizing and understanding these 
factors, but from that point on the responsibility of marketers is that of strategy, 
which is more like a practical engineering job: actually refracting and focusing the 
"light" to get maximum reaction of the desired kind. 



COMPETITIVE POSITION 361 



ROLE OF BRAND 



The principal way to focus appeals is through use of a brand— sometimes the 
manufacturer's own name, as in the case of Ford automobiles; more often a dis- 
tinct and, hopefully, distinctive name created for the purpose, as in the case of 
Lucky Strike cigarettes (made by the American Tobacco Company); but always 
a communication means whereby distributors, salesmen, and consumers may 
readily identify the product. 

Brand is a focusing device because it is a symbol— whether in letters or other 
graphic forms, whether visual or oral, and whether or not it is trade-marked (as it 
usually is, if it is distinctive, mnemonic, and qualifies for legal protection against 
being imitated). As a symbol, it can communicate all the different appeals that 
the product has for all kinds of people. Once a product means something to peo- 
ple, the brand name evokes the feeling that led to its being bought before, and 
that resulted from its being used . . . and leads to its being bought again. Many 
companies feel that the brand recognition and loyalty they have built up over a 
long period of promotion is worth millions of dollars. 

How Many Brands? 

But this brings up an intrinsic question: Should a manufacturer or store- 
keeper take advantage of the existing feeling about a brand and try to transfer it to 
other products, or should he keep it concentrated on one product? If he uses a 
"family" brand, like Campbell, covering a whole family of different soups, or 
Heinz, covering a family of many different kinds of products (more than 57!), 
or Ann Page, covering a wide variety of products sold in A & P stores, then he 
can do an intense job of promotion on the one brand. On the other hand, some- 
times he may make more total sales by having a number of specialized brands — 
like the Procter & Gamble soap company's Ivory, Camay, Tide, Rinso, Gleem, 
Crest, and so on. The answer depends on a number of factors: 

( 1 ) The transferability of the brand. This depends in large part on whether 
the products in question are bought and used in the same way. Thus, those 
who like a brand of chicken soup will tend also to like the same brand of 
mushroom soup; but those who like a brand of tomato soup will not tend to 
the same degree to like the same brand of tomato ketchup. 

(2) The amount of money that can be spent on promotion. If the com- 
pany can afford to do a full-scale job of promotion on a number of brands, it 
will have a sharper focus for each; but if it can't, then it may be better off by 
sacrificing some individual sharpness for the sake of the more diffuse focus 
but heavier impact it can achieve through putting all of its resources in back 
of one brand. 



362 COMPETITIVE POSITION 

(3) The degree of market saturation secured by existing brands. When a 
company has, say, 60 per cent of the total market with one brand (A), and 
finds it difficult or too expensive to increase that share against a competing 
brand (B), it can sometimes make headway with a new brand having a new 
appeal (C). Even if it takes away 5 per cent of A's existing market, it will 
put the company ahead as long as it also gains something from C's market. 
Thus: 

Before: A, 60%; B, 40% 

After: A, 55% plus C, 10% = 65%; B, 35% 

The situation can be repeated again and again, as B responds with its new 
brand D, and A and C are supplemented with E, and so on. 

All three of these considerations hinge on the fact that a brand can develop 
a particular bundle of appeals for a particular group of people, which, for these 
people is stronger in its effect than a wider, more varied set of appeals. Sometimes, 
piling appeals on top of appeals actually cuts the number of people being com- 
municated to. For example, as reported by Fred T. Schreier, drawing on his ex- 
periences as Director of Research of A. J. Wood Research Corporation: 

In a recent study that we made of driver preferences for gasoline stations, 
we found two features or characteristics which seemed to have high motiva- 
tional power. If the driver thought that a specific station was likely to give 
"quick service," there was a fairly high probability that the driver would stop 
at that station if he were on the road. We also found that if the driver felt 
that there was "lots of equipment" at a station, there was likewise a fairly 
high probability that the driver would stop there if he were on the road and 
in need of gas. 

However, if he felt both that the station was the kind that would give 
"quick service" and that it was a station with "lots of equipment," then the 
probability was lower than for either of the two features separately. 

An exploration into the reasons for this behavior indicated that if all the 
driver wanted was to get in and out of the station quickly, he would be likely 
to choose a "quick service" station. But if what the driver wanted was to have 
great care and attention given to his car, there was a good likelihood that he 
would choose an "equipment" station. But those stations which were char- 
acterized by both of these features satisfied very few drivers— the feeling being 
that, where there was lots of equipment, one could not really expect quick serv- 
ice; and where there was the appearance of quick service, there was less likeli- 
hood of great care in spite of the equipment. 4 

On the other hand, if the appeals are not conflicting, then a number of them 
will increase the probabilities of finding a variety of compatible existing predispo- 



4 Fred T. Schreier, "Seven Fallacies in Marketing Logic," Harvard Business Review, Sep- 
tember-October 1959, p. 116. 



COMPETITIVE POSITION 363 

sitions, of securing exposure and attention, and of bringing that many more people 
closer to buying. There may well be a considerable amount of opportunity cost in 
assuming that for practical purposes everyone is more or less alike. 5 This is par- 
ticularly true in today's world with its great number of people able to buy almost 
everything they want; a segment of the market may be almost as great as the 
whole market used to be in the past. Thus, the sales of the Falcon, which is just 
one of Ford's range of models, are now almost as great as total Ford sales 15 years 
ago. Such segments with different needs and wants are large enough to support 
both large-scale production and large-scale promotion. (See the fuller discussion of 
market segmentation for product policy, in Section Six.) 

There is always some overlap between the segments covered by two or more 
appeals. As a result, there is some gain: the effect of two appeals on the same people 
can be greater, since they now have a number of paths to the same goal. But there 
also is some offsetting loss: the effect of two appeals can evoke conflict, either as 
described in the service station example above, or because people sometimes find 
it more difficult to make up their minds when they have several ways to do the 
same thing. Thus, in direct-mail selling, a dual offer of 1 year for $6.00, 2 years for 
$10.00 will usually sell fewer subscriptions than just 1 year for $6.00 (although the 
total dollar return may be greater). The net result is that two appeals usually have 
more total power than one, but not twice as much— and the exact amount will 
depend on how much of the overlap is reinforcing, and how much of it is 
conflicting. 



Optimum Combination of Appeals 

In all situations, the most efficient strategy will be the one where (a) the re- 
sult of number of people likely to receive the message times (b) the attractive 
power of the appeal is the greatest— it can be many people and a low-power appeal, 
or few people and a high-power appeal. Or the situation can be more complex as 
shown in Table 5-D, which shows the relative effectiveness of three appeals in 
varying combinations, and using one, two, or three brands. 

The reasoning is simple, but the estimating is difficult. In essence, the per- 
centage of the market responsive to the various combinations of appeals and 
brands is multiplied by the attractive power of the particular combination (usually 
greater for a larger number, but not proportionately greater, as explained above), 
to yield the potential market share; and the resulting figure is then multiplied by 
a cost index, representing the relative economic efficiency of that combination 
(that is, with 1.0 representing the efficiency of a single appeal, what it will cost to 
get the same rate of efficiency with the combined appeals and brands), to yield an 
index of relative effectiveness. 



5 Donald F. Cox, "Looking Around: Clues for Advertising Strategists: II," Harvard Business 
Review, November-December 1961, pp. 180-182. 



TABLE 5-D. CALCULATION OF OPTIMUM COMBINATION OF DIFFERENT APPEALS 

AND DIFFERENT BRANDS 



Part One. Evaluation of Appeals 







// 






III 




/ 


Estimated per cent of 


Estimated 




Appeal (s) 


market responsive 


to 


appeal (s) 


power of appeal (s) 


1. 


A only 


10% 






5% 


2. 


B only 


5% 






20% 


3. 


C only 


5% 






40% 


4. 


A&B 


35% 






.75 a (.05 + .20) = 19% 


5. 


A&C 


25% 






.25 b (.05 + .40) =11% 


6. 


B&C 


10% 






.50 c (.20 + .40) =30% 


7. 


A&B&C 


10% 






.20 d (.05 + .20 + .40) = 16% 



100% 



Part Two. Evaluation of Courses of Action 







V 
Calculated 












cumulative 


VI 










per cent 


Calculated 


VII 


VIII 






of market 


potential 


Estimated 


Index of 




IV 


responsive 


market 


cost 


effectiveness 




Courses of Action 


to appeal (s) 


share 


index 


Col. VI -r Col. VII x 10 


8. 


One brand, one appeal, A 


80% e 


4% f 


1.0 


0.4 


9. 


One brand, one appeal, B 


60% 


12% 


1.0 


1.2 


10. 


One brand, one appeal, C 


50% 


20% 


1.0 


2.0 


11. 


One brand, two appeals, 
A&B, combined 


95%* 


16% h 


1.2 1 


1.3 


12. 


One brand two appeals, 
A&C, combined 


95% 


18% 


1.2 


1.5 


13. 


One brand, two appeals, 
B&C, combined 


90% 


17% 


1.2 


1.4 


14. 


One brand, three appeals, 
A, B&C, combined 


100% 3 


19% k 


1.6 1 


1.2 


15. 


Two brands, using A&B 
separately 


95% m 


16% n 


1.8° 


0.9 


16. 


Two brands, using A&C 
separately 


95% 


24% 


1.8 


1.3 


17. 


Two brands, using B&C 
separately 


90% 


32% 


1.8 


1.8 


18. 


Two brands, using A&B 
combined, C separately 


100% p 


36%* 


2.1 r 


1.7 


19. 


Two brands, using A&C 
combined, B separately 


100% 


30% 


2.1 


1.4 


20. 


Two brands, using B&C 


100% 


21% 


2.1 


1.0 




combined, A separately 










21. 


Three brands, using A, 
B&C separately 


100% 


36% s 


2.6* 


1.4 



a This .75 multiplier represents the estimate that the two appeals, A and B, supplement 
each other (but, because of overlap, not fully), hence have a total power well above the 
average of their individual powers. Thus, the estimated power of A only is 5%; of B only, 
20%; but of A and B combined, .75 times the sum of 5% and 20%, or'l9%. (The straight 
average would have been .50 times the sum of 5% plus 20%, or 13%.) 

b This .25 multiplier represents the estimate that the two appeals, A and C, actually de- 
tract from each other, hence have a total power considerably less than the average of their 
individual powers. 

364 



COMPETITIVE POSITION 365 

Notes to Table 5-D, Continued 

c This .50 multiplier represents the estimate that the two appeals, B and C, neither sup- 
plement nor detract from each other. 

d This .20 multiplier represents the estimate that the three appeals, A, B, and C, have a 
total power a little less than the average of their individual powers. 

e For example, this 80% is the sum of the Column II values for all items in Column I 
containing an A. Similar calculations apply to the figures in Column V, Rows 9 and 10. 

f For example, this 4% is the product of Column V, Row 8 times Column III, Row 1. 
Similar calculations apply to the figures in Column VI, Rows 9 and 10. 

g For example, this 95% is the sum of the Column II values for all items in Column I 
containing either an A or B. Similar calculations apply to the figures in Column V, Rows 12 
and 13. 

h For example, this 16% is the sum of the products of Column II times Column III for 
all items in Column I containing either an A or B. (Thus, A & B has power in the segment 
of the market shown as being responsive to A & C, Row 5, because both combinations 
have A in common. Here, the power is figured as that of the combined appeals A & C, 
rather than just the A element. The reasoning for using this figure is: (a) the range of ap- 
peals to which a person is responsive also represents his range of interests; (b) his interests 
will tend to become fully engaged since competition sees to it that whatever applicable 
appeals are available will get used by someone; and (c) his susceptibility to any combination 
of appeals will therefore be in proportion to the appeals that he is responsive to rather than 
in proportion to the appeals that any one company or product happens to offer him.) 
Similar calculations apply to the figures in Column VI, Rows 12 and 13. 

1 This 1 .2 index represents the estimate that doubling of appeals necessitates more space 
(time) and/or frequency for full effect, but not double (there is considerable economy). 

J This 100% is the sum of all values in Column II. 

k This 19% is the sum of the products of Column II times Column III for all items in 
Column I. (See footnote h for explanation of use of Column II figures.) 

x This 1.6 index represents the estimate that tripling of appeals necessitates more space 
(time) and/or frequency for full effect, but not triple (there is considerable economy). 

m This 95% is the same as Column V, Row 11. Similar observations apply to the figures 
in Column V, Rows 16 and 17. 

n For example, this 16% is the sum of Column VI, Row 8 plus Column VI, Row 9. 
Similar calculations apply to Column VI, Rows 16 and 17. 

This 1.8 index represents the estimate that using two brands about doubles cost (there 
are minor savings) . 

p This 100% is the same as Column V, Row 14. The identical observation applies to the 
figures in Column V, Rows 19, 20, and 21. 

« For example, this 36% is the sum of Column VI, Row 11 plus Column VI, Row 10. 
Similar calculations apply to the figures in Column VI, Rows 19 and 20. 

r This 2.1 index represents the estimate that using an extra appeal necessitates more space 
(time) and/or frequency, but not proportionately (there is considerable economy). 

s This 36% is the sum of Column VI, Row 8; plus Column VI, Row 9; plus Column VI, 
Row 10. 

l This 2.6 index represents the estimate that using three brands about triples cost (there 
are minor savings ) . 

A number of interesting implications develop from such an analysis: 

(1) An appeal which is likely to draw a strong response from a small 
percentage of the market, C, is, in this illustration (it might be quite different 
in another product-market situation), a better bet than an appeal which is 
likely to draw a weak response from a large percentage of the market, A. 

(2) Appeals A and B are reinforcing, B and C don't do anything for or 
against each other, but A and C definitely detract from each other— as might 



366 COMPETITIVE POSITION 

be the case, for example, if A were an appeal based on economy, B an appeal 
based on convenience or service, and C an appeal based on high fashion. 

(3) The analysis indicates the sequence of courses of action which (as- 
suming that enough effort can be put into any one of the alternatives to reach 
the minimum necessary for a practical effect) a marketer should follow. In 
this particular case he would follow this order: (a) a single brand, with one 
clearly defined appeal (Row 10 in Table 5-D); then, what that had secured 
as much of the market as it could be expected to, the addition of another 
brand with a clearly defined appeal of its own (Row 17); and finally, (c) when 
both previous moves had reached a saturation or near-saturation point, the 
addition of another appeal to one of the brands (Row 18). This move would 
be particularly appropriate in view of the fact that adding a new appeal to an 
already established brand will usually cost less than starting a new brand. 

(4) These steps would have to be fitted into an over-all strategy, which 
would include long-range market-share objectives. 

Thus, Dr. Charles C. Slater, Consulting Economist, Arthur D. Little, Inc., 
shows how a manufacturing company, with its own retailing operations, will fare 
under four different 5-year market share objectives. Table 5-E applies to mature 
industries where demand is relatively stable and to a great degree independent of 
the number of outlets in the market; both food retailing and gasoline retailing 
meet these criteria. The calculations show that: 

If the firm holds 20 per cent of the market and acts on a program to lose 
1 per cent of its market a year, it is likely to make a good profit on the retail- 
ing end of the business. In contrast, an aggressive campaign to gain at the 
rate of 1 per cent a year would sharply reduce retailing profits. On the other 
hand, profits from the manufacturing and warehousing activities are illus- 

TABLE 5-E. PROFIT EXPECTATIONS FROM RETAILING AND FROM WAREHOUSING 
AND MANUFACTURING UNDER FOUR DIFFERENT MARKET SHARE OBJECTIVES 

(in thousands of dollars) 



•5 iS 



t; £•£> b£ "Sii &:§ §£--£ ^ £ 5 * 

ill 111 ?$| St 11 |ll HI |.l! 

■li jfc III II If lit 111 111* S83 

A -1.0 $ 950 $190 $ 95 $58 $37 $30 $67 

B - .5 975 195 98 63 35 37 $72 

C 1,000 200 100 70 30 44 $74 

D +1.0 1,050 205 105 95 10 51 $61 



Charles C. Slater, "The Most Profitable Market Share Objectives," Journal of Marketing, April 1961, 
p. 55. 



COMPETITIVE POSITION 367 

trated as having risen in this latter case. A stable market share sales rate is 
the most profitable rate of effort; that is, market share objective C. 6 
Again, what we have just seen is only a framework for thinking, no better 
than the thought that goes into it as a pattern for logical reasoning, and no bet- 
ter than the thought that goes into the estimates that are its raw material. It 
should stimulate thought rather than be a substitute for it. 

THE QUESTION-ANSWER PROCESS 

The point is: Can we get the information we need to make the estimates 
with which to do such systematic, analytical, scientific thinking? Can we uncover 
and interpret the attitudes and opinions of people? Inevitably, here, we come face- 
to-face with the difficulties and opportunities in marketing research as a way, not 
just of counting noses, not just of counting feelings and ideas, either, but as a way 
of perceiving and measuring subtle differences among attitudes and opinions, and 
their development over time. 

We have already looked, in Section Three, at some of the problems involved 
in deciding (a) whom to question, (b) what medium to use to reach them, and 
(c) how many to go after so as to have the necessary number of responses. Now 
let us look at the problem of how to question them. Of all the possible ways in 
which truth can get distorted, the manner of questioning may be the most im- 
portant—not just because mistakes in the area are inherently crucial, but also 
because they are temptingly easy to overlook. 

Questions can be so leading (unintentionally) that they become misleading; 
thus, if you ask people to tell you how much they like a product, you are almost 
forcing them to say they like it. Or answers can be misleading, too, because they 
are superficial or partial, and the questioner takes them at their face value. There 
is need both to guard against inadequate questioning and to positively plan how 
to get the information your test calls for. 

There are two issues to decide here: (a) how openly to ask the questions, and 
(b) how explicitly to guide the answers. 

The danger in asking questions in direct form is that people may tend to 
answer more conservatively or sophisticatedly than they feel, either because they 
are trying to live up to what is expected of them (by the questioner or by them- 
selves) or because they actually do not know how they feel. Just asking an open 
question may cause them to give more attention to a product, or think more self- 
consciously about it, than they would in real life— which can bring about a dan- 
gerously exaggerated effect, particularly when one is considering a new product, 
since even seeing (or becoming conscious of) the product may be more than most 
of the respondents would have done in a store where the product was on sale. 

So care must be taken, in the questioning, to help the respondent answer as 
accurately as he or she wants to. 



6 Charles C. Slater, "The Most Profitable Market Share Objectives," Journal of Marketing, 
April 1961, p. 55. 



368 COMPETITIVE POSITION 

Granted, there are situations where a direct, open question is the simplest 
approach— especially when emotional or subjective feelings are less involved, as 
with purchasing agents' reactions to technical products. But in more complex 
situations it is absolutely necessary to take one or more of the following courses. 
(The first two, pertaining to ease and interest, also are useful in connection with 
direct, open questions.) 

Ease of Answering 

The easier the answering, the more effective the questioning will be. This 
does not mean that complicated or lengthy questionnaires or interviews must be 
avoided— in fact, respondents are usually willing to work harder than most ques- 
tioners expect them to— but simply that, for any given set of questions, you should 
try to make them as easy to answer as possible. Here there are a number of prin- 
ciples to keep in mind: 

a. In any event, make the beginning questions easy— something the re- 
spondent can answer without "half trying," such as "Do you watch TV?" 
(Incidentally, you can use this kind of opening to screen out people you 
don't want to get answers from in the first place, and simply use the oppor- 
tunity to ask the sometimes very valuable question, "Why not?") 

b. Try to keep the order natural and logical— keeping related questions 
together and with one question leading to another, from the more general to 
the more specific, in this kind of order: 

1. Do you watch TV? 

2. During what periods of the day? 

3. What programs? 

4. Which is your favorite? 

5. Why? 

c. On individual questions involving effort to remember or estimate, you 
must not ask far too much accuracy or people will be reluctant to answer. 
For example, "Approximately how many times did you try Brand X during 
the last month?" is better than asking for an exact number; while ". . . during 
the past week?" would make it even easier. Sometimes it is helpful to offer 
general categories for checking (either on a questionnaire or on a card handed 
to the respondent in an interview), such as: 

Not at all □ From 5 to 10 times □ 

Less than 5 times □ Over 10 times □ 

In general, of course, it is easier for people to check a box— Yes Q No □ 
—than to actually write in "yes" or "no"; or to say "yes" or "no" than to an- 
swer in descriptive phrases like "twice a week" or "I prefer Brand X"; or to 
circle an item in a list than to name their favorite brand. And plenty of 
"white space" on the questionnaire, or a leisurely manner in the interview, 



COMPETITIVE POSITION 369 

will make it seem shorter than trying to crowd it into a fewer number of 
pages or a fewer number of minutes. 

Often, of course, you must compromise between your need for more de- 
tail and your knowledge that the more you ask for, the less accuracy you will 
get. Of particular importance is whether you allow for alternative answers like 
"Don't know," "Don't care," "Other," when asking specific responses to 
questions such as "What brand of toothbrush do you use?" or "What brand 
of sugar do you prefer?" or "Which of these magazines do you read?" 

On difficult questions, if you want to know whether people actually do 
know or not, giving the respondents the alternative of not reallv answering 
specifically will be revealing, as well as easier. On the other hand, if you want 
to force people to indicate which way they lean in their opinions or memories, 
and are willing to pay the price of discouraging respondents a little, then you 
will not give them an easy way out, like "don't know." 

d. Keeping the respondent interested, which is discussed as the next 
general principle, also must be mentioned here because the effect of interest 
is always to make difficult things seem more easy. 

Creating and Maintaining Interest 

Here there are two possibilities (or opportunities) to create and maintain 
interest: (1) in the subject of the questioning, and (2) in the process of the ques- 
tioning. 

The former has already been discussed to some extent in Section Three; it de- 
pends mostly on the approach to the respondent, that is, the letter accompanying 
the questionnaire (or the introduction at the top of the questionnaire) or the 
opening statement in the interview. This element may actually be the biggest 
determinant of the success of the questioning; it not only creates and maintains 
interest, but it also produces alertness on the part of the respondent. 

As for the process of questioning, you can "buy" people's interest by offering 
a reward for those who cooperate, and/or you can achieve interest by (1) the 
variety of the forms in which you ask questions (yes-and-no questions, open-end 
requests for comments, grids for checking (as in Case 4-9), items to rank, and so 
forth); and (2) by the sequence of the questions, so that, as with the problem of 
ease, the respondent naturally moves from one query to another, and therefore his 
original interest mounts rather than flags. 

Casual or Buried Questions 

In order to elicit answers that are not colored by the fact that the respondent 
gives them more attention than he or she would in real life, the crucial question 
has to be presented casually. Thus, the stated subject of the questioning may be 
made broader than the issue which is the objective of the research; for example, if 
the purpose is to find out about preferences for canned soup, the inquiry may be 



370 COMPETITIVE POSITION 

built around cooking or eating habits, and the questions about soup do not stand 
out as being important. 

A further variation of this approach is actually to bury or "hide" the question 
by making it one item among many. Thus, in word-association techniques (such 
as described in Case 2-6), the product whose reputation is the focus of the ques- 
tioning may be one word in a list of 15 or 20; or, in an image study, the company 
may be one name among 4 or 5 which respondents are asked to rate on various 
characteristics like progressive-conservative, scientific-commercial, warm-cold, and 
so on (so-called "semantic differentials"). 

This also gives some valuable opportunity for making use of comparison. The 
other items among which the significant one is buried should, to the extent that 
this does not impair the appearance of reasonableness in the question, be chosen for 
what they will yield in the way of differences. Thus, a medium-size company whose 
image is being tested might find it useful to know how it stacked up against a 
large company, a small company, as well as several companies its own size (in- 
cluding, perhaps, one it already considered was similar to it and one it suspected 
might be very different). 

Exaggeration 

Another way to bring out opinions, particularly those that people do not 
realize they have or that involve such small differences in attitude they would have 
difficulty in expressing them, is to exaggerate the conditions of the test. This can 
be done in three ways: 

(1) Within the question itself— The principle can be illustrated by the 
researcher asking people to choose between a very sour drink and a very sweet 
drink when all he is interested in is the effect of a small degree of change in 
his product. 

(2) Within the way the question is administered— This can be illustrated 
by the researcher giving people more time to appraise an ad than they would 
usually accord it in real life, or measuring reactions to prominent displays 
of a product in a store, whereas normally people would only see it less con- 
spicuously on the shelves. (Just the fact that in store tests the shelves are 
usually kept filled, so as to avoid the operation of another variable in the way 
the shelves look, also has some exaggerating effect.) 

(3) Within the way the answers are analyzed— This can be illustrated 
by the researcher who forces a large sample of people to choose between dif- 
ferent combinations of similarly desirable or similarly undesirable adjectives 
(courageous-progressive, ingenious-forward looking, cowardly-dishonest, and so 
forth) and then uses an electronic computer to search for possible patterns 
of significance. In one such case, it was revealed that executives had more of a 
leaning toward weak subordinates than they themselves realized. 7 



7 Lewis B. Ward, "Problems in Review: Do You Want a Weak Subordinate?" Harvard 
Business Review, September-October 1961, p. 6. 



COMPETITIVE POSITION 371 

An example which involves several of these principles is that of the researcher 
who wanted to find out whether more motorists would be attracted by a sign that 
was a symbol (a flying red horse) or a word (MOBIL). (At the time, the flying 
red horse was the main identifying sign of Mobilgas stations.) Using the research 
hypothesis that the attractive power of a sign would be in proportion to the degree 
to which it was seen and recalled, he made this test: 

An automobile equipped with a movie camera was driven along roads 
where Mobilgas and other gas stations were located. As the camera moved 
along, it photographed every gasoline outlet on the way. After the film was 
developed, the number of Mobil stations and other brands was equalized. To 
avoid bias, the selection was random. 

Various audiences were invited to see the movie thus prepared, and after- 
wards every member in each audience was requested to tell what brands of 
station signs were on the road, and approximately how many of each he 
thought there were. The frequency of the flying red horse sign was under- 
estimated down to about one-half of that of its most efficient competitor. 8 

Despite its very ingenious simulation, this case also exemplifies the purposeful 
use of exaggeration in the form of the question posed, in the way it was posed, and 
in the analysis of results. The issue is how far to go in exaggerating without chang- 
ing the nature of the results rather than just the degree. A researcher, wishing to 
get evidence for deciding between a pastel blue and a pastel yellow package, might 
get a completely misleading answer if he used a deeper blue and a more intense 
yellow in his test because people who are attracted by the pastel yellow may not 
like a more intense yellow at all. Similarly, people who will switch to frozen soups 
when they are just as cheap as canned soups may hardly switch at all for anything 
less, so exaggerating the price difference as a way of trying to find out what 
would happen in the case of a small reduction again might be misleading. 

Another problem is the fact that sometimes the exaggeration is unintentional 
and is not allowed for in the interpretation of the results. The fact remains that, 
if used purposefully, which also means carefully planned, exaggeration can be a 
useful tool. 



Experimental Games 

If the test can be given the feeling of a game, people will "let themselves go" 
and in the process reveal their own feelings. Professor Edgar A. Pessemier of the 
State College of Washington describes an experiment in which 103 subjects went 
on a series of simulated shopping trips, with the goal being to maximize the satis- 
faction that could be obtained from a combination of merchandise and money. 
Each participant was told how much money he had available to spend, the assort- 
ment of brands available in each class of goods for which he was "shopping," and 



8 Alfred Politz, "The Dilemma of Creative Advertising," Journal of Marketing, October 
1960, p. 2. 



372 



COMPETITIVE POSITION 



the price of each item. The means by which he could maximize his satisfaction 
were defined by the funds and assortments available on each shopping visit. 

One of the results is pictured in Figure 5-E; notice that people apparently 
have less "brand loyalty" to toothpaste than to cigarettes— they switched tooth- 
paste brands in larger numbers and as a consequence of less of an increase in the 
price of their preferred brand. 

Pessemier recognizes that the emphasis on price was somewhat exaggerated 
by listing the items and their prices in a manner that encouraged comparison. 
(This would of course serve to bring out differences, of the kind shown in Figure 
5-F, that might otherwise be hard to detect; the need for caution applies to trying 
to measure the size of the difference.) However, Pessemier utilized an interesting 
device which both added reality and yet kept the price effect from being as ex- 
aggerated as it might otherwise have been: after each experiment one "shopper," 
selected on a random basis, was given the actual items and change called for by 
the selections made during one of his simulated shopping trips. Therefore, the 
subjects really tried to maximize their satisfactions in case they were a "winner," 
but nowhere nearly so compulsively as if they knew they were going to end up for 
sure with the actual products they bought and the real change they had left. 9 



90 







— 


80 





Brand Switching ^ 


„, -•- *"" ■"" 


70 









60 




Tooth Paste jr'" 


— 


50 





f Cigarettes 





40 


_ 


/ s 


__ 


30 
20 
10 


Si* 


/ / 

*l 1 1 1 1 


No. of No. of 
Buyers Brands — 

Tooth Paste 102 7 
Cigarettes 48 9 

1 1 1 1 



100 
90 
80 
70 
60 
50 
40 
30 
20 
10 



3 4 5 6 7 

BRAND PRICE INCREASE IN CENTS 



10 



Figure 5-F. Relative Brand Loyalty in Experimental Game Showing Cumulative Percentages of 
Buyers Who Switched Brands as Their Preferred Brands Increased in Price 



9 Edgar A. Pesseimer, "A New Way to Determine Buying Decisions," Journal of Marketing, 
October 1959, pp. 41-43. 



COMPETITIVE POSITION 373 



Role-Playing 

Another game-like approach, but without the action and simulated reality of 
an experimental game, is to ask respondents what they think someone else's opin- 
ion or reaction will be. The someone else can be an anonymous person in a picture 
or cartoon; can be an "average housewife" or "most people" in a question; or can 
be the object of a matching process whereby people are instructed to indicate which 
of a list of "lawyers, truck drivers, bankers, machinists, professional athletes, farm- 
ers, college students, and so on" are most likely to drink "Brand X beer, Brand Y 
beer, Brand Z beer, and so on." 

One company using this kind of test found that, through its advertising, it 
had unintentionally created an image of being the beer of lower-class people. Since 
respondents were classified by various social-class criteria and had also been asked 
to state their brand preferences for beer in another part of the test, some further 
analysis was possible. And it was found that the image did not help much to 
sell it even to lower-class people, who liked the idea of "trading up," in the sense 
of buying what they considered a higher-quality product, yet seriously hurt its 
appeal to the rest of the population. 

The whole idea in back of such role-playing questions is that people, in de- 
scribing others' opinions and reactions, will reveal their own, free of inhibitions 
imposed by having to take a position or by having to confess something about 
themselves. 

Figure 5-G shows another matching of roles to brands. The image of Brand 
A is more austere, less friendly; the consumers see it not only as of higher status 
but also somewhat remote. On the other hand, Brand B has a warmer sound to it, 
down-to-earth quality; it is regarded as a more ordinary brand, one for everyday 
people. 

Particularly where the various brands of a product are largely indistinguish- 
able as to quality, cost, effectiveness, and so on— or, what is about the same thing, 
each user group claims that its brand is outstanding on the same points— it takes 
techniques like this (and incomplete sentences, word association, forced choices, 
semantic differentials, and so on) to bring out what consumers have difficulty in 
stating— or that they even deny. 10 



Indirect and Disguised Questions 

Role-playing, just described, is a form of indirect questioning, too— that is, 
getting an answer to a question by an indirect or circuitous route. 

Both the answer and the question can be indirect. Thus, when a female re- 
spondent checks an age category on a card handed to her, she is answering indi- 



10 Burleigh B. Gardner and Sidney J. Levy, 'The Product and the Brand/' Harvard Business 
Review, March -April 1955, p. 36. 



374 COMPETITIVE POSITION 

PER CENT 
90 




Figure 5-G 

rectly, and somehow finds this easier than answering directly. However, the oppor- 
tunity for indirection is greater with questions than with answers. Here are a 
number of possibilities, ranging from the very simple to the complex: 

(a) Intended question: How frequently do you buy product x? 
Question asked: When did you last buy product x? 

(The question asked is easier; and if there are enough people in the 
sample, the rate of frequency can be deduced from the answers.) 

(b) Intended question: What brand of product x do you prefer? 
Question asked: What brand of product x do you have in your refrigera- 
tor? 

(The question asked will be answered more accurately; and while spe- 
cial deals and couponing may cause another brand to be in the refrig- 
erator at the moment of interviewing, the total measurement may be a 
very definite reflection of preference.) 

(c) Intended question: How old are you? 

Question asked: How long have you lived in this city (or town)? 
The question itself will produce answers that provide a fair indication 
of the average age of respondents; but if in addition you ask a question 
elsewhere in the questionnaire or interview as to "Have you lived here 
all your life?" and "If not, when did you move here?" then you really 
get an accurate picture.) 



COMPETITIVE POSITION 375 

(d) Intended question: What kind of executive are you? 

Question (s) asked: What mazagines do you read? What organizations 
and associations do you belong to? What degrees do you have? 
(The researcher must infer the answers he wants from the answers he 
gets.) 

(e) Intended question: How important do you consider this feature of prod- 
uct x? 

Question asked: How much more would you pay for a product with this 
feature? 

(If you can get people to put a monetary value, no matter how ar- 
bitrary, on different features of a product, you can get an idea of their 
feeling about the relative value of such features.) 

(f) Intended question: Do you consider the advertisement or the salesman 
a more reliable source of information about new products? 

Question asked: In this situation how do you rate the purchasing agent 
pictured ... [in the context of a critical situation: for example, a break- 
down of an important piece of machinery, for one group of respond- 
ents; in the context of a routine situation, for another, different 
group]? 

(If, in addition, the two groups of respondents are each broken into two 
subgroups, one of which is shown a picture of a purchasing agent look- 
ing at an advertisement, and the other of which is shown a purchasing 
agent talking to a salesman, considerable insight can be gained by see- 
ing which subgroup or subgroups rate the purchasing agent as most 
professional, businesslike, and so forth. Presumably, the more critical 
situation will cause a switch toward the more highly regarded source 
of information.) 
The essence of this and other indirect questioning is to infer opinions or atti- 
tudes from related phenomena (often with the aid of elaborate computer calcula- 
tions). Obviously, there is both considerable richness of opportunity and danger 
of miscalculation. Accordingly, it is essential to confirm the relationship in further 
tests whenever the decision to be made on the basis of the findings is important. 
Sometimes, some fairly direct follow-up is possible when the findings have pin- 
pointed the crucial attitudes or opinions, including the most direct method of all- 
testing sales in the market. 

Spontaneous Reaction 

Finally, there is the kind of questioning where the questioner simply tries to 
get the respondent to talk spontaneously about his or her feelings. Some stimulus 
in the form of questions or pictures is needed, but they should be nonleading ques- 
tions or neutral-looking pictures (in contrast to the use of exaggeration, where they 
might be deliberately provoking). 



376 COMPETITIVE POSITION 

This method requires considerable skill and depends for its success on careful, 
as well as intuitive, analysis. Obviously, it is the least quantifiable. 



STRUCTURING THE ANSWERS 

We have, in discussing how openly to ask questions, anticipated many of the 
points involved in the other important aspect of the questioning-answer process: 
how definitely to guide or structure the answers, by devices requiring the respond- 
ent to say "yes" or "no," or "I prefer this adjective to that one," or "I rate this 
factor 1, 2, 3, 4, or 5," or "I agree with this statement," and so on, rather than 
simply letting the respondent react spontaneously. 

The factors which determine just how far to go in guiding or structuring 
need further exploration, however. In general, what you do in this connection 
depends on what you are seeking. Thus: 

1. Degree of precision— for example, asking people to say, not just whether 
they like something or not, but whether they like it very much, like it a little, 
don't care one way or the other, dislike it a little, or dislike it very much; or 
asking people to rank their preferences for a number of products, adjectives, 
statements, or whatever. 

The danger is that if you ask people to be more precise than they really 
can be, then, provided you don't discourage them from answering at all, you 
invite them to answer dishonestly or carelessly, and thus defeat your whole 
purpose of wanting precision in the first place. 

As a matter of fact, if you give people 3 or 5 gradations of opinions to 
check or rank, you will usually get the same general results in terms of 
over-all favor or disfavor, agreement or disagreement, as you would if you 
gave them 7 or 9 to check or rank, or, for extreme precision, asked them to 
make the requisite number of paired comparisons (that is, in the case of 9 
items, you could ask them to choose between a and b, a and c, a and d, and so 
on, through the 36 possible pair combinations). Also, if you ask people to 
name the one product (or whatever it may be) that they like most out of a 
list, the total single votes for the different items will usually be in the same 
order as the results you would get from a ranking question with the votes 
weighted. 

On the other hand, it is possible to draw more detailed inferences from 
answers to the more detailed form of questions, particularly if you want to 
watch what happens to the attitude of the same people before and after some 
marketing move. 

2. Stretching for effect— {ox example, deliberately asking people for more 
than they know or feel they can answer honestly or intelligently, for the sake 
of getting them to reveal their attitudes and opinions (even though they 



COMPETITIVE POSITION 377 

may not be conscious of how they feel). This is, of course, one of the indirect 
devices mentioned earlier. Thus, if you ask people to choose between two 
adjectives, or say yes or no, where they ordinarily wouldn't, you make them 
"tip their hand." 

This, again, can result in some irresponsible answers, yet if you do it 
purposely, with your eyes open, you often can get a total finding that is re- 
vealing of how the majority of your respondents actually feel. Thus, when 75 
per cent of respondents choose "professional" over "scientific" in describing 
the information they get about drugs from one pharmaceutical manufacturer, 
and 60 per cent make the same choice in the case of another manufacturer, it 
can be inferred that the former manufacturer has a higher connotation of 
professionalism than the latter. 

A good example of how structuring can have a stretching effect is the 
kind of test where you ask respondents to .indicate their agreement or dis- 
agreement with various statements, and purposely arrange the series of state- 
ments so that (a) in general, the order is such that it will both draw people 
out and identify different shades of opinion; and (b) specifically, the first 
question is so extreme that nobody is expected to agree, but it "takes the 
edge off" the next statement so that, even though people would ordinarily not 
admit (to themselves or the tester) that they agreed, they will be more willing 
to let themselves go and indicate the agreement that they actually feel. For 
instance, the first question might be, "All housewives are so lazy they don't 
deserve husbands"; the second, "The majority of housewives prefer not to 
wash dishes the moment a meal is over." 

3. Aid to remembering— for example, listing items for people to check, or 
mentioning items for them to react to, rather than asking, "What brands 
do you buy?" or "What magazines do you read?" or "What advertisement do 
you recall?" or whatever. 

The effect of aiding memory is to avoid deflating the position of less well- 
known, less frequently purchased or seen, or less prestigeful items, at the risk 
possibly of inflating it. (There are ways, to be discussed shortly, for verifying 
answers. The risk of inflation is likely to be easier to take care of than the risk 
of deflation, because it is easier to spot errors of commission, as for instance, 
people saying they read a magazine when they don't, than errors of omission, 
as, for instance, people forgetting to mention a magazine they do read.) 

An example of deflation is the case of an appliance manufacturer who was 
third in industry sales with a 10 per cent share of market, but who always 
showed up with only 2 per cent or 3 per cent of consumer preferences; except 
for people who had just bought or were about to buy, most people auto- 
matically said they preferred one or the other of the two leading brands that 
together had 90% of the market and were advertised about 10 times as heavily. 

On the other hand, if the purpose of the test is to find out which brand, 



378 COMPETITIVE POSITION 

magazine, or ad comes to mind more readily, as an indication of the relative 
power of the promotion or established image of the various competitive 
items, then unaided recall can be a more fruitful method (as long as it isn't 
taken as a measure of actual buying or preference). As a matter of fact, in 
deciding between the aided and unaided methods, marketers may find it bet- 
ter to use whichever one is more like the actual buying situation; that is, if the 
sale depends on people spontaneously asking or going after a particular prod- 
uct, then the unaided approach may be a more appropriate indication of rela- 
tive competitive position; but if people are presented, by display in the store 
or by the salesman, with a series of alternatives, then the aided approach 
would seem more suitable. 

One other situation where the unaided method would be called for is 
where the tester himself does not know what items are available or exposed 
to the respondent. 

4. Ease of administration— iox example, asking questions in such a way 
that the answers can be readily interpreted, classified (coded, is the technical 
term), and analyzed; and/or so that results can be verified and quantified. 

On the first point, all that needs to be said is that no structuring at all is 
most difficult; that a certain amount of structuring (particularly if pretested, 
and the code numbers for punched cards or other method of tabulation al- 
ready established— even to the extent of being included on the questionnaire 
itself) is most convenient and least expensive; and that a large amount of 
detailed structuring can lead to excessive time and cost (there are many 
projects where, say, $5,000 worth of computer analysis takes place just be- 
cause the material is there and it looks so "juicy," but the results are worth 
only a little more than, say, $1,000 or $2,000 worth). 

The second point, relating to the fact that structuring facilitates verifica- 
tion and quantification, needs further exploration (see below). 
It should be clear by now that there are many ways in which even subtle and 
deep-seated attitudes can be brought out in enough definite form to be (a) verified 
and (b) quantified. How far to go in structuring, then, becomes a question of 
whether one knows enough about the objectives of the test to plan the structuring 
intelligently— as compared with catching some unexpected information by less 
definitive, more "grab-bag" techniques. 

Verification 

Getting verification within the question-answer process takes a number of 
forms, ranging from the simple to the complex, as follows: 

FACTUAL VERJFJCATJON-asking any who indicate they prefer or 
regularly use brand x pens to bring one to the door in exchange for a new 
one. Also, total (or average) answers to questions such as automobile owner- 



COMPETITIVE POSITION 379 

ship can be compared with known or published statistics; presumably, if the 
particular answers are correct, then the other answers can be depended on, 
too. This can be handled along with checking the characteristics of the ef- 
fective sample, to see if it is representative, as noted in Section Three. (Ex- 
perience indicates that putting an identifying number or other marks on the 
questionnaire or return envelope for this purpose does not have a serious de- 
terring effect on returns.) 

INTERNAL CONSISTENCY— arranging the questionnaire or inter- 
view, or other test, so people will have to answer the same question or react 
to the same stimulus a number of times in different form; consistent reactions 
can be taken as an indication (but not proof) of the true aim of the question- 
ing, that is, the actual opinion of the respondent. 

Another device is to deliberately try to trip up the respondent with a 
false question. For example, in readership studies, the respondents can be 
asked whether they have read a list of articles, with one or two imaginary titles 
included, and then the answers can be deflated by the degree of "imagina- 
tion" that the respondents show. 

It should be emphasized that most inaccuracy is not the result of de- 
liberate dishonesty on the part of respondents— they just find it easier to think 
about certain things in certain ways. Among other things they usually find it 
easier to check the first or the last name on a list; or they get tired toward 
the end of a long questioning process; or they slip into a routine of answering 
in the same pattern. There are ways to avoid such effects, as by rotating names 
or questions— but the best way, because it raises the level of all answers to all 
questions, is to keep the respondent alert by keeping him interested. (People 
of higher education or position are less prone to diluting effects, perhaps be- 
cause they take a more positive interest in the first place.) 

STATISTICAL SIGNIFICANCE-testing the differences between the 
results of tests included at different times or on different samples, or between 
the results to different questions on the same test. 

The simplest way to do this is by use of the formula, 



-< 



2 1 (Q_\2 



+ (S 



where S d is the standard deviation of the difference, and S a and S b are the 
standard deviations of two test findings (computed as shown in Section Three) . 
S d is then divided into the difference between the two test findings, and the 
resulting figure shows what the chances are (or how much confidence you 
can have) that the difference did not occur by chance; in other words, that 
it is a real difference. 

Thus, suppose last year you found from a test where the effective sample 
was 100 that 55 per cent of the respondents said they preferred your brand, and 



380 COMPETITIVE POSITION 

this year with an effective sample of 200 you get 60 per cent preference. Then: 



= ^.003675 = .06 = 6% 

Next, that 6 per cent is divided into the observed difference, 60 per cent 
—55 per cent = 5 per cent. The result, 5 per cent -^- 6 per cent = .83, is less 
than 1 standard deviation. The changes then are less than 68 out of 100 that 
the difference is real; or, to put it the other way, the probability of such a 
difference occurring by chance, even though no real difference existed, is 
more than 32 out of 100. Or, again, you can have less than 68 per cent con- 
fidence. 

On the other hand, take this situation (for variety, let's test the dif- 
ference between two findings from the same survey) : with a sample of 300 
you find that 50 per cent prefer your Brand A, 40 per cent prefer Brand B, 
and 10 per cent prefer Brand C; and you want to know how much confidence 
you can have that the difference between your brand's preference and B's is 
real rather than the result of chance. Then (this time simplifying the equation 
by eliminating the square root signs of S a and S b , since S a and S b need to be 
squared anyway) : 

Sd = \j -5° • - 50 + - 40 ' - 60 = J .00083 + .0008 
v 300 300 \ 



-V 



.00163 = .04 = 4% 



Next, when 4 per cent is divided into the observed difference, 10 per cent, 
the result is 2 Vi— between 2 and 3 standard errors and therefore around 98 
per cent, or (using a statistical table of the "areas under the normal curve" 
and multiplying by 2 to get both sides of the bell) 98.758 per cent. That is, 
there are less than 2 chances out of 100 that there is an unreal difference that 
showed up through "bad luck" in the sample; or, again, we can have almost 
99 per cent confidence that the difference is real as far as the effect of sample 
error is concerned. (It could be completely wrong if the questioning had been 
misleading, however. ) 

Another useful way of testing differences is the Chi-Square method, which 
will produce the same results as the previous method on two findings, but is 
more versatile in that it allows more than two differences to be tested at the 
same time. It would be possible, for example, to test the differences among 
Brands A, B, and C simultaneously, and also to test several dimensions of 
preference like strongly prefer, prefer, don't know, strongly dislike. When the 
various findings make a pattern that is internally consistent, the resulting con- 
fidence figure is higher than for the difference between the individual findings. 

For example, Professor Robert Ferber of the University of Illinois, in 



COMPETITIVE POSITION 



381 



Statistical Techniques in Market Research, gives an example resulting from a 
survey of 913 people on their plans to purchase life insurance. (See Figure 
5-H.) The Chi-Square method works (we shall not go into its details here, 
since it is explained in standard statistical texts, like Ferber's) on the prin- 
ciple of finding the differences between the observed findings (for example, 
17 in the upper left-hand cell, 174 in the lower right-hand cell) and what 
would be the expected figures for the corresponding cells if the distribution 
of the different categories had occurred by chance (for example, 110/913 X 
202 = 24 in the upper left-hand cell, 237/913 X 619 = 161 in the lower right- 
hand cell), and then testing cumulatively; the other observed differences are 
equally as small as the two just specified, but the total confidence figure is 
over 99 per cent. 

MARKET FOR LIFE INSURANCE BY INCOME 
LEVEL 



Income Level 

Under $2,000 
$2,000-$2,999 
$3,000-$4,999 
$5,000 and over 
Total 



(Sample of 913) 

Plan to 

purchase 



Undecided 



Not planning 
to purchase 



17 


23 


70 


56 


23 


177 


87 


25 


198 


42 


21 


174 



202 



92 



619 



Total 

110 
256 
310 
237 
913 



Source: Robert Ferber, Statistical Techniques in Marketing Research (New 
York, McGraw-Hill Book Company, 1949), p. 268. 

Figure 5-H 



Again, it must be emphasized that this kind of figure testing only helps to 
eliminate (or warn about) the possibility of random error making differences 
look real that actually are not. There is just as much danger, if not more, in 
failing to plan the questioning carefully. 

COMMON-SENSE ANALYSIS. One further-and not inconsiderable 
—way to verify findings is to see if they make sense when they are analyzed. 
There are many times when the refusal of a figure to fit into a pattern that 
"hangs together" internally and jibes with judgment and experience has called 
attention to the fact that somewhere in the process of collecting and coding 
the information a serious error was made: some interviewer misunderstood the 
instructions, some coder developed a whopping blind spot, or some question 
was structured in a way to reverse the intended meaning. 

This is a better test to find how much confidence not to have in the re- 
sults than how much to have. But if the researcher has thought his problems 
through analytically, and pored over the questionnaires or tapes or interview 
reports himself, alert for patterns that no one ever thought to tell the com- 



382 COMPETITIVE POSITION 

puter to produce, he often can make some positive finds, also. In particular, 
it is worthwhile to experiment with classifying the information in different 
ways— for example, first (a) by occupation and under occupation, by income, 
then (b) by income and under income, by occupation, and so on— to see 
which way a more meaningful pattern will result. 



Quantification 

Finally, the combination of probability techniques in sampling, psychological 
research devices in questioning, and multiple correlation in analysis (including 
manual searching for patterns of cause and effect when a computer is not available 
or the economic value of the project does not justify the expense of computer 
work) makes it increasingly possible for marketing research to be scientific. (Vari- 
ance analysis, regression analysis, and so on, offer many types of statistical tools 
not mentioned here.) 

It is possible to collect large masses of opinions expressed in such form as 
Agree Q Neutral □, Disagree □, or Very Important Q, Important Q Indiffer- 
ent □, Unimportant □, Very Unimportant □; and all the other ways referred to 
in previous discussion or illustrated in cases, plus classification data about the 
respondents. Then, by manipulating the various factors, holding first this one con- 
stant and then another (in hundreds or thousands of possible patterns, if you 
want to, on a computer), significant relationship can be established— such as that, 
although income and education are highly correlated, in certain product or service 
areas, education may have more of a determining effect on consumer acceptance 
of innovations than does income, whereas income may have more bearing in con- 
nection with price resistance. 

In a very real sense, the manipulation of data demonstrates one of the major 
aspects of scientific marketing: controlled experimentation. In this event, you can 
do the experimentation after the material has been collected, by putting the data 
together in different forms and figuring, for example, what would result if you 
could concentrate on particular kinds of people with particular kinds of prefer- 
ences, responsive to particular kinds of appeals, who exist in a known ratio to 
other people. (Naturally, it is better to plan the collection of data with definite 
purposes in mind, but there always is room for further experimentation.) Thus: 

A cosmetics manufacturer, who had been spending large amounts of 
money trying desperately to increase his 5 per cent share of market among 
younger women, but to no avail, might find that the particular appeal which 
won a ready acceptance for his brand among older women was alienating the 
younger ones, and further, that his advertising message had still not gotten 
through to more than 30 per cent of the older women. So, even though the 
younger market was very important in total potential, he should switch the 



COMPETITIVE POSITION 383 

aim of his advertising to the older market where he was already better estab- 
lished and had more of a chance of increasing sales substantially. 11 
Here would be a good case for considering whether he would not gain by intro- 
ducing a new brand at this point. That is, rather than simply trying to expand 
sales of his existing products, perhaps he should reconsider and plan his whole 
product strategy, as discussed in the next section. 

* * * 

Whatever the share of market a company has, management seeks to take 
action to increase it; the various ways of developing selective demand are illustrated 
in the following cases, either in the form of proposed marketing moves or of 
appraisals of existing competitive strength. 



11 Adapted from Fred T. Schreier, op. cit., pp. 112-113. 



384 



COMPETITIVE POSITION 



In the Musselman Department Store 
(located in Mill town and catering to 
the low-to-medium income group) David 
Dirkson, salesman for the Zenith Drug 
Company (a wholesale drug firm) is 
calling on Beatrice Bishop, the buyer for 
the Cosmetics and Drug Sundries De- 
partment. 

Salesman: Miss Bishop, I represent the 
Zenith Drug Company. We're 
expanding our services to offer 
a wholesale prescription serv- 
ice, which I'm sure your cus- 
tomers would find attractive. 

Buyer: A wholesale prescription serv- 
ice? I've never heard of that. 

Salesman: It is novel, isn't it? — and 
that's why it should mean 
some new, plus business for 
you. 

Buyer: But what is it? How does it 
work? 

Salesman: When your customers are here 
in the store, shopping for 
other items, no doubt right 
here in your department, they 
leave their prescriptions with 
you, and we pick them up at 
5:00 p.m., fill them over- 
night, and have the medicine 
all neatly put up in bottles 
and labeled — with the store's 
name, Musselman Department 
Store Prescription Service — 
back at 9:00 a.m. the next 
morning, ready for delivery. 

Buyer: But where is the advantage? 

Salesman: Your customers save 25 per 
cent of the retail price, 25 per 
cent off what they'd pay in a 
drugstore. There is a lot of 



CASE 5-1 

Musselman Department Store 

talk these days about the high 
price of drugs. 25 per cent — 
that's a lot of money on some 
of the new miracle drugs that 
cost $5.00 or $10.00 a pre- 
scription. 

Buyer: That does sound interesting. 
Tell me more about your firm. 

Salesman: Oh, we've been in business in 
Exeter County for over 75 
years, and now we're operat- 
ing here in Monroe County — 
have been for several months 
now. As a matter of fact, 
we've chosen Monroe County 
as the scene for this new pre- 
scription service. 

Buyer: Why? 

Salesman: Well, partly because we don't 
have as many retail drugstore 
accounts here; they might not 
welcome this particular kind 
of competition. And also we 
think Monroe is a good coun- 
ty. And Milltown, particularly, 
is a real shopping center. 

Buyer: What kind of markup would 
we get? 

Salesman: If you want to give your cus- 
tomers the full 25 per cent 
saving, you'd still get the tra- 
ditional department store fig- 
ure of 40 per cent of the sell- 
ing price; or you could fatten 
your share by giving them a 
saving of only 15 or 20 per 
cent. 

Buyer: Do you think it would fit in 
with cosmetics, and vitamins, 
and bathing caps, and so on? 
We don't carry any medicines 
now, you know. 



COMPETITIVE POSITION 



385 



Salesman: Of course it would. And it What is the problem underlying the 



might be the entering wedge 
for a full-scale drug business. 
The regular drugstores carry 
enough department store mer- 
chandise, Lord knows! 



decision Miss Bishop must make? What 
are the questions she must ask herself — 
and answer? Should she take on the new 
service? Why would Musselman gain 
competitively by this move? Why would 
Zenith Drug gain? How good a salesman 
is Dirkson? 



386 



COMPETITIVE POSITION 



CASE 5-2 

Cantwell Packing Company (A) 



In September, 1956, Mr. Harold Mc- 
Gowan, vice president in charge of sales 
of the Cantwell Packing Company, flew 
from Syracuse, New York, where the 
company was located, to Boston, Massa- 
chusetts, to hear a presentation concern- 
ing the proposed fall-winter advertising 
budget by the account executive of the 
company's advertising agency, Hamilton 
and Calston. The Cantwell company 
packed tomato ketchup, tomato juice 
cocktail, sliced pickled beets, red cab- 
bage, stewed tomatoes, and succotash. 
These products were sold throughout the 
New England area. 

In 1955 the company's sales totaled 
$2,040,000. Cost of goods sold had been 
$1,628,000, and sales expense and gen- 
eral and administrative expense had to- 
taled $282,000, leaving a before-tax profit 
of $130,000. The results of 1955 very 
closely approximated results of the pre- 
ceding four years. The total 1955 adver- 
tising budget had been $75,000 — evenly 
divided between the spring-summer and 
the fall-winter campaigns. Cantwell had 
retained Hamilton and Calston for a 
number of years. 

Part of the agency's presentation, "Rec- 
ommendations on Advertising for Cant- 
well Products, October, 1956 to March 
31, 1957/' which was accompanied by 
illustrative slides, follows: 



Plan 

We have a plan for six months of ad- 
vertising and sales promotion in New 
England. The planning has had our most 
careful consideration. We believe it to be 
sound in every respect and that it will be 
worth what it costs. 



Whenever we submit a plan for adver- 
tising and sales promotion, we feel that 
our client has every right to ask: "What 
will this advertising and promotion do 
for me?" "What will I get back for the 
money I spend?" 

Since advertising and selling are never 
a matter of plain arithmetic, we are not 
able to say, "Invest X dollars in this plan 
and you will get Y dollars back." Never- 
theless, the question can be answered, if 
the planning is sound, on a basis of rea- 
sonable probabilities of what benefit will 
result. This benefit can be evaluated. 

In order to make such evaluations, the 
campaign must have specific aims. The 
probability of accomplishing these aims 
must be judged and a value set on the 
accomplishment. 

Objectives 

The plan has three aims: 

(a) To gain new consumers. 

(b) To sell more Cantwell products 
to present consumers. 

(c) To move all Cantwell products. 
The first stated aim is the big objec- 
tive. The other two aims are corollaries. 



Sales Potentials 

The first question we should face when 
thinking about added Cantwell custom- 
ers in New England, or adding sales by 
whatever means, is the size of the market. 
Is there room in this market for more 
Cantwell sales? Is the market saturated? 
Is competition so strong that no further 
progress can be made? 

Following are the National Canners 
Association figures of the U.S. pack of 



COMPETITIVE POSITION 



387 



items in which you are interested. The 
figures do not match similar figures put 
out by the U.S. Department of Com- 
merce which, apparently, include some 
operations not recognized by the Na- 
tional Canners Association: 



pack of ketchup, we find that New Eng- 
land probably purchased 1,128,000 cases. 
Your sales constitute but a small part of 
this total. No brand of ketchup is so 
strongly entrenched in this territory that 
it makes a successful bid for new busi- 



Product 



U.S. ANNUAL PACK 
(millions of standard cases) 



J 950 



1951 



1952 



1953 



1954 



Tomato Juice 


22.7 


31.6 


31.4 


32.7 


23.4 


Ketchup 


14.8 


24.0 


19.6 


19.1 


18.5 


Beets (all types) 


8.1 


7.5 


6.7 


8.6 


7.0 


Tomatoes 


18.7 


27.6 


25.1 


20.8 


20.9 


Succotash 


0.241 


0.259 


0.318 


0.488 


0.31 



In 1950, The Bureau of Industrial Sur- 
veys of the American Newspaper Pub- 
lishers Association made a panel survey 
among housewives on food products. It 
shows the market by areas of the nation 
and the following figures cover tomato 
products: 



ness by Cantwell in any way impossible. 
Following a similar process of reason- 
ing, the present annual canned beet mar- 
ket for New England is estimated at 
430,000 cases. Under the Cantwell brand 
you supply relatively only a small part of 
this demand, and no competitor is suffi- 



PER CENT SALES DISTRIBUTION BY AREAS 





New 


Middle 


East 






West 






England 


East 


Central 


Central 


South 


Central 


Pacific 


Per Cent 
















U.S. Population 


6.5 


15.4 


31.9 


8.4 


18.2 


9.0 


10.6 


Per Cent 
















U.S. Sales 
















Tomato Sauce 


2.0 


37.7 


9.2 


5.4 


11.5 


3.7 


30.5 


Tomato Ketchup 


6.1 


18.8 


29.9 


10.5 


12.9 


11.4 


10.4 


Chili Sauce 


6.4 


15.0 


35.2 


13.7 


7.3 


6.6 


15.8 


Paste and Puree 


4.9 


39.6 


15.5 


6.7 


10.1 


2.9 


20.3 



For example, let's take ketchup. New 
England contains 6.5 per cent of the U.S. 
population and accounts for 6.1 per cent 
of ketchup sales. In short, it is just 
slightly on the minus side of being a 
completely normal market for this prod- 
uct. It is way under average, however, for 
tomato sauce and somewhat under nor- 
mal for paste and puree. 

If we apply these figures to the 1955 



ciently dominant to interdict the success 
of your effort for a larger share of the 
market. 

On tomato juice (we have no break- 
down for tomato cocktail) the normal 
New England annual market figures out 
as 1,490,000 cases — certainly big enough 
and not too competitive to warrant a try 
for more business for Cantwell. 

The picture is much the same for 



388 



COMPETITIVE POSITION 



canned tomatoes, with a New England 
potential of 1,279,000 cases. 

The figures for succotash show a total 
estimated present New England consump- 
tion of 18,400 cases. 

No figures as to pack or potential seem 
to be available on red cabbage. 

The local representative of a national 
packer has told us that he believed that 
there was good opportunity for sales in- 
crease in the New England market for 
stewed tomatoes; that his company's sales 
are slowly